Comparison of physical and performance characteristics of NCAA Division I football players: 1987 and 2000

The purpose of this study was to compare normative data from present Division I National Collegiate Athletic Association football teams to those from 1987. Players were divided into 8 positions for comparisons: quarterbacks (QB), running backs (RB), receivers (WR), tight ends (TE), offensive linemen (OL), defensive linemen (DL), linebackers (LB), and defensive backs (DB). Comparisons included height, body mass, bench press and squat strength, vertical jump, vertical jump power, 40-yd-dash speed, and body composition. Independent t-tests were used to analyze the data with level of significance set at p < 0.01. Significant differences (p < 0.01) were found in 50 of 88 comparisons. From 1987 until 2000, Division I college football players in general have become bigger, stronger, faster, and more powerful. Further research is warranted to investigate if these trends will continue.     

Head impact exposure in collegiate football players

In American football, impacts to the helmet and the resulting head accelerations are the primary cause of concussion injury and potentially chronic brain injury. The purpose of this study was to quantify exposures to impacts to the head (frequency, location and magnitude) for individual collegiate football players and to investigate differences in head impact exposure by player position. A total of 314 players were enrolled at three institutions and 286,636 head impacts were recorded over three seasons. The 95th percentile peak linear and rotational acceleration and HITsp (a composite severity measure) were 62.7g, 4378rad/s(2) and 32.6, respectively. These exposure measures as well as the frequency of impacts varied significantly by player position and by helmet impact location. Running backs (RB) and quarter backs (QB) received the greatest magnitude head impacts, while defensive line (DL), offensive line (OL) and line backers (LB) received the most frequent head impacts (more than twice as many than any other position). Impacts to the top of the helmet had the lowest peak rotational acceleration (2387rad/s(2)), but the greatest peak linear acceleration (72.4g), and were the least frequent of all locations (13.7%) among all positions. OL and QB had the highest (49.2%) and the lowest (23.7%) frequency, respectively, of front impacts. QB received the greatest magnitude (70.8g and 5428rad/s(2)) and the most frequent (44% and 38.9%) impacts to the back of the helmet. This study quantified head impact exposure in collegiate football, providing data that is critical to advancing the understanding of the biomechanics of concussive injuries and sub-concussive head impacts.     

The effects of training history, player position, and body composition on exercise performance in collegiate football players

Performance data for 261 NCAA Division 1A collegiate football players were analyzed to determine if player position, body weight, body fat, and training time were correlated with changes in performance in the following events: power clean (PC), bench press (BP), squat (SQ), vertical jump (VJ), 40-yd dash (40yd), and 20-yd shuttle (20yd). Individual positions were combined into the following groups: (A) wide receivers, defensive backs, and running backs, (B) linebackers, kickers, tight ends, quarterbacks, and specialists, and (C) linemen. Increases in body weight were positively correlated with increases in BP and PC performance for all groups. Increases in body fat were negatively correlated with performance in the PC and VJ for all groups. For group C, increases in body fat were also negatively correlated with performance in the 40yd and 20yd. Group and training time exhibited no linear relationship with performance in any of the tested events. No linear relationships were observed between the independent variables and performance in the SQ. When individual training data were analyzed longitudinally, a nonlinear increase in performance in the PC, BP, and SQ was observed as training time increased, with the greatest rate of change occurring between the first and second semesters of training.     

Equipment and running surface alter sprint performance of college football players

The purpose of this study was to determine the effect of football equipment and running surface on sprint performance in NCAA Division II football players (n = 68). Players were timed in the 40-yd sprint on an indoor rubberized track (Day 1) and on an outdoor, natural-grass football field (Day 2) wearing either regulation football equipment or shorts and a T-shirt. Each player was assigned randomly to perform 2 trials under each condition on each surface, and the average of the 2 trials was used for analysis. Offensive backs, defensive backs, and linebackers were significantly faster than were offensive and defensive linemen in all trials, and subjects were collapsed into 2 groups, backs and linemen. Football equipment significantly impaired performance on the track (-2.8% +/- 1.7%) and the field (-2.9% +/- 1.8%). The increase in body mass due to the football equipment was significantly greater for backs (7.2% +/- 0.7%) than for linemen (6.5% +/- 1.0%), but produced a significantly greater impairment in sprint performance in linemen (-3.3% +/- 1.1%) as compared with backs (-2.5% +/- 1.5%). Sprint performance was significantly and equivalently impaired when running on grass (backs: -2.5 +/- 1.1%; linemen: -2.8 +/- 1.4%) as compared with the track. Thus, running a 40-yd sprint in football equipment on a natural grass field impairs performance by an average of 5.5% (+/- 2.3%) compared with running indoors with minimal apparel. Football equipment and running surface significantly impair sprint performance in college football players, the effect being greater in linemen than in backs, and is likely related to differences in muscle strength/power and body fat.     

Metabolic and thermoregulatory responses to a simulated American football practice in the heat

Energy cost is a major factor influencing the tolerable thermal load, particularly during exercise in the heat. However, no data exist on the metabolic cost of football practice, although a value of 35% of maximal aerobic capacity (VO(2)max) has been estimated. The energy cost and thermoregulatory response of offensive linemen (OL) was measured wearing different American football ensembles during a simulated half of football practice in the heat. Five collegiate offensive linemen (133 kg, 20% fat, 42 ml x kg(-1) x min(-1) maximal oxygen uptake) completed each of four 60-minute test sessions in an environmental chamber (28 degrees C, 55% relative humidity [RH]) wearing shorts (S), helmet (H), helmet and shoulder pads (HS), and full gear (FUL). Core temperature in the digestive tract (TGI) was obtained using an ingestible sensor. During simulated football drills (e.g., repetitions of drive blocking), exercise intensity ranged from 30 to 81% VO(2)max but averaged 55%VO(2)max (6.7 METS) overall. Blood lactate remained >5 mmol x L(-1), and heart rate (HR) averaged 79%HRmax. Equipment had a significant effect on %VO(2)max but only during recovery between drills with HS (61.4 +/- 3.7%) compared with H (53.3 +/- 6.9%) and S (40.1 +/- 8.5%). The TGI was higher (p < 0.05) with HS compared with H at several time-points after 30 minutes. Football practice for OL elicits a significantly higher overall metabolic cost (>6 METS, >50%VO(2)max) than assumed in previous studies. The addition of shoulder pads increases core temperature and energy cost, especially during recovery between active drills in unacclimatized linemen.     

Relationship between football playing ability and selected performance measures

The relationships between football playing ability (FPA) and selected anthropometric and performance measures were determined among NCAA Division I-A football players (N = 40). Football playing ability (determined by the average of coaches' rankings) was significantly correlated with vertical jump (VJ) in all groups (offense, defense, and position groups of wide receiver-defensive back, offensive linemen-defensive linemen, and running back-tight end-linebacker). Eleven of 50 correlations (groups by variables), or 22%, were important for FPA. Five of the 11 relationships were related to VJ. Forward stepwise regression equations for each group explained over half of the criterion variable, FPA, as indicated by the R(2) values for each model. Vertical jump was the prime predictor variable in the equations for all groups. The findings of this study are discussed in relation to the specificity hypothesis. Strength and conditioning programs that facilitate the capacity for football players to develop forceful and rapid concentric action through plantar flexion of the ankle, as well as extension of the knee and hip, may be highly profitable.     

Physical and performance characteristics of Japanese division 1 collegiate football players

Iguchi, J, Yamada, Y, Ando, S, Fujisawa, Y, Hojo, T, Nishimura, K, Kuzuhara, K, Yuasa, Y, and Ichihashi, N. Physical and performance characteristics of Japanese division 1 collegiate football players. J Strength Cond Res 25(12): 3368-3377, 2011-This study aimed to establish the physical and performance characteristics of football players in the Japanese Division 1 collegiate football program and perform a comparison of these characteristics between Japanese (n = 208) and US Division 1 football players (n = 797). The following comparisons were made: (a) between a higher-ranked university team vs. a lower-ranked university team in Japan, (b) between different playing positions in Japan, (c) between starters and nonstarters in Japan, and (d) between playing positions in Japan vs. those in the United States. The results of this study suggest that players in the higher-ranked university team were heavier, stronger in back squat, jumped higher, and had greater power than those on the lower-ranked team. Furthermore, linemen were generally characterized by larger size, greater strength, and more fat as compared with backs. On the other hand, backs tended to be faster, smaller in physical size, have higher vertical jump height, and show greater relative strength than linemen did. Starters were taller, heavier, stronger, had more powerful, and more fat-free mass than nonstarters. Finally, our results revealed that players in the United States were superior to players in Japan in all body status comparisons (p < 0.01). This study revealed that performance and superior body composition are essential for the success of a football player. Power and strength seem to be key factors in defining good football performance.     

Body composition in Division I football players

This study assessed body composition of Division I football players (n = 69) and compared the findings with previously reported data to ascertain whether the increase in player total body mass that has been observed over the past 10 years has been accompanied by an increase in body fat. Body composition was determined by hydrostatic weighing and the measurement of skinfold thicknesses. Total body mass, skinfold thicknesses, and body fat were greater in the current players than in players in studies conducted in the early 1980s and early 1990s. Body fat varied significantly across playing position, with the defensive backs, offensive backs, and receivers being the leanest and the offensive linemen and tight ends the most fat. There was no significant relationship between body composition and playing year or scholarship status, nor were any differences observed between ethnic groups. Of important clinical relevance was the finding that the linemen (offensive, defensive) and tight ends were on average greater than 25% body fat, the borderline for obesity in this age group. Much of this fat was deposited in the abdominal region, a significant finding when one considers the high correlation between abdominal obesity and ischemic heart disease and stroke. The current findings suggest that more attention needs to be given to the nature of the increase in body mass being achieved by today's football player to minimize long-term negative health consequences, and the findings reemphasize the need identified in earlier studies of the importance of detraining programs for these athletes.     

Comparison of Olympic vs. traditional power lifting training programs in football players

Twenty members of an National Collegiate Athletic Association Division III collegiate football team were assigned to either an Olympic lifting (OL) group or power lifting (PL) group. Each group was matched by position and trained 4-days.wk(-1) for 15 weeks. Testing consisted of field tests to evaluate strength (1RM squat and bench press), 40-yard sprint, agility, vertical jump height (VJ), and vertical jump power (VJP). No significant pre- to posttraining differences were observed in 1RM bench press, 40-yard sprint, agility, VJ or in VJP in either group. Significant improvements were seen in 1RM squat in both the OL and PL groups. After log10-transformation, OL were observed to have a significantly greater improvement in Delta VJ than PL. Despite an 18% greater improvement in 1RM squat (p > 0.05), and a twofold greater improvement (p > 0.05) in 40-yard sprint time by OL, no further significant group differences were seen. Results suggest that OL can provide a significant advantage over PL in vertical jump performance changes.     

The National Football League combine: a reliable predictor of draft status?

The performance of 326 collegiate football players attending the 2000 National Football League combine was studied to determine whether draft status could be predicted from performance measurements. The combine measured height and weight along with 9 performance tests: 225-lb bench press test, 10-yd dash, 20-yd dash, 40-yd dash, 20-yd proagility shuttle, 60-yd shuttle, 3-cone drill, broad jump, and vertical jump. Prediction equations were generated for 7 position categories with varying degrees of accuracy-running backs (RBs), r(2) = 1.00; wide receivers (WRs), r(2) = 1.00; offensive linemen, r(2) = 0.70; defensive linemen, r(2) = 0.59; defensive backs (DBs), r(2) = 1.00; linebackers, r(2) = 0.22; and quarterbacks, r(2) = 0.84. The successes of the prediction equations are related to the ability of the individual tests to assess the necessary skills for each position. This study concludes that the combine can be used to accurately predict draft status of RBs, WRs, and DBs. The equations can also be used as a good to fair estimate for other positions.     

Body composition and physical performance in men's soccer: a study of a National Collegiate Athletic Association Division I team

The purpose of this study was to examine the relationship between body composition (BC) and physical performance (PP) in male collegiate soccer players and differences among positions and between starters and non-starters. Twenty-seven male collegiate soccer players were tested at the beginning of the 2003-2004 season (age = 19.9 +/- 1.3 years, height = 177.6 +/- 6.3 cm, body mass = 77.5 +/- 9.2 kg, body fat (BF) = 10.6 +/- 5.8 kg, and %BF = 13.9 +/- 5.8%). BC, vertical jump (VJ), speed (S), lower-body and total body power production (TPW), and estimated Vo(2)max were measured. Values found for BC were similar than the ones in the literature. Significant correlations were found between BC and PP ranging from -0.38 to 0.61 for weight, VJ, S, TPW, and Vo(2)max. BF showed a positive correlation with S (r = 0.60) and a negative correlation with Vo(2)max (r = -0.67). The values for BC and PP were similar in starters and non-starters with only TPW showing a significantly greater value in starters. It is apparent that all members of a team train to play owing to the long seasons and substitutions, and a high level of excellence is demanded of both starters and non-starters alike. Training programs that equally benefit both groups are important in soccer.     

Pulmonary O2 uptake kinetics as a determinant of high-intensity exercise tolerance in humans

Tolerance to high-intensity constant-power (P) exercise is well described by a hyperbola with two parameters: a curvature constant (W') and power asymptote termed "critical power" (CP). Since the ability to sustain exercise is closely related to the ability to meet the ATP demand in a steady state, we reasoned that pulmonary O(2) uptake (Vo(2)) kinetics would relate to the P-tolerable duration (t(lim)) parameters. We hypothesized that 1) the fundamental time constant (τVo(2)) would relate inversely to CP; and 2) the slow-component magnitude (ΔVo(2sc)) would relate directly to W'. Fourteen healthy men performed cycle ergometry protocols to the limit of tolerance: 1) an incremental ramp test; 2) a series of constant-P tests to determine Vo(2max), CP, and W'; and 3) repeated constant-P tests (WR(6)) normalized to a 6 min t(lim) for τVo(2) and ΔVo(2sc) estimation. The WR(6) t(lim) averaged 365 ± 16 s, and Vo(2max) (4.18 ± 0.49 l/min) was achieved in every case. CP (range: 171-294 W) was inversely correlated with τVo(2) (18-38 s; R(2) = 0.90), and W' (12.8-29.9 kJ) was directly correlated with ΔVo(2sc) (0.42-0.96 l/min; R(2) = 0.76). These findings support the notions that 1) rapid Vo(2) adaptation at exercise onset allows a steady state to be achieved at higher work rates compared with when Vo(2) kinetics are slower; and 2) exercise exceeding this limit initiates a "fatigue cascade" linking W' to a progressive increase in the O(2) cost of power production (Vo(2sc)), which, if continued, results in attainment of Vo(2max) and exercise intolerance. Collectively, these data implicate Vo(2) kinetics as a key determinant of high-intensity exercise tolerance in humans.     

Prediction of 1 repetition maximum in high-school power lifters

Eighteen elite male power lifters performed 1-repetition maximum (1RM) and submaximal strength tests (70, 80, and 90% 1RM) to develop prediction equations for the squat (SQ), bench press (BP), and deadlift (DL) exercises. For each equation, stepwise multiple-regression prediction procedure included the maximum number of repetitions (REPS) completed at a given %1RM weight (REPWT). For SQ and BP the 70% 1RM yielded the best 1RM prediction equations: (1RM SQ [kg]) = 159.9 + (0.103 x REPS x REPWT) + (-11.552 x REPS), with a standard error of the estimate (SEE) of 5.06 kg; (1RM BP [kg]) = 90.66 + (0.085 x REPS x REPWT) + (-5.306 x REPS), with an SEE of 2.69 kg. For DL the 80% 1RM yielded the best prediction equation: (1RM DL [kg]) = 156.08 + (0.098 x REPS x REPWT) + (-12.106 x REPS), with an SEE of 4.97 kg. The athlete's years lifted (number of years of power lifting experience) was highly correlated with the 1RM strength for BP and DL (r > 0.70) but not for SQ (r < 0.70). No bodily structural dimension variable had a significant correlation with 1RM strength (r < 0.70). The results of this study indicate that 1RM SQ, BP, and DL may be predicted with an acceptable degree of accuracy in elite male high-school power lifter subjects.     

The accuracy of the American College of Sports Medicine metabolic equation for walking at altitude and higher-grade conditions

The purpose of this study was to examine the accuracy of the American College of Sports Medicine (ACSM) walking equation at low walking speeds, altitude (1,550 m), and higher grades. Twenty men and women (mean +/- SD, age, 28 +/- 6 years; height, 171 +/- 13 cm; weight, 67.8 +/- 18.1 kg) completed 2 randomized testing sessions under altitude (AL) (P(I)o(2) = 123.1 mm Hg [20.93%]) and sea level control (SLC) (P(I)o(2) = 147.3 mm Hg [25.00%]) conditions. Steady-state oxygen uptake (Vo(2)) was measured while subjects walked at 50 m.min(-1) at 8 separate grades (0, 5, 10, 15, 18, 21, 24, and 27%). Steady-state Vo(2) measurements from the last 2 minutes of each grade in AL and SLC were compared to the predicted Vo(2) of each grade according to the ACSM walking equation. Mean Vo(2) differences between predicted and AL values ranged from -0.5 to 1.4 ml.kg(-1).min(-1), averaged -0.1 ml.kg(-1).min(-1) across all grades, and were significant (p < 0.05) at 0 percent grade. Mean Vo(2) differences between predicted and SLC values ranged from 0.6 to 3.0 ml.kg(-1).min(-1), averaged 1.4 ml.kg(-1).min(-1) across all grades, and were statistically significant (p < 0.05) at 0 and 5 percent. The standard error of the estimate (SEE) for the prediction of Vo(2) under AL and SLC were 2.2 and 2.0 ml.kg(-1).min(-1), respectively. Total errors for the prediction of Vo(2)max under AL and SLC were 2.3 and 2.6 ml.kg(-1).min(-1), respectively. Overall, the findings indicate that the current ACSM prediction equation for walking is appropriate for application at low speeds, moderate altitude, and higher grades.     

Identification of a Vo2 deflection point coinciding with the heart rate deflection point and ventilatory threshold in cycling

The purposes of this study were to compare the patterns of the work rate (WR)-Vo2 and WR-heart rate (HR) relationships in incremental cycling, to ascertain the occurrence of a Vo2 deflection (Vo2def) coinciding with the HR deflection point (HRdef ), and to determine whether the Vo2def, if present, coincides with the ventilatory anaerobic threshold (VT). Twenty-four professional cyclists performed a maximal incremental test on a wind-load cycle ergometer. Work rate, HR, Vo2, and Vco2 were recorded. The WR-Vo2 relationships obtained were linear up to submaximal WR and curvilinear thereafter and thus described a Vo2def. The WR and Vo2 at Vo2def were mathematically determined for all subjects. The ratio of DeltaWR.DeltaVo2 up to Vo2def was significantly lower than that above Vo2def (90 +/- 11 W.L.min versus 133 +/- 35 W.L.min, p < 0.0001). The WR-HR relationships obtained were linear up to submaximal WR and curvilinear thereafter. The WR and HR at HRdef were mathematically determined for all subjects. The WR values at Vo2def and at HRdef (329 +/- 32 W and 326 +/- 34 W) were significantly correlated (R = 0.96, p < 0.0001) and in good concordance (limits of agreement from -4.7% to 3.2%, Bland-Altman analysis). The Vo2 at VT was then determined for all subjects. The Vo2 values at Vo2def and at VT were significantly correlated (R = 0.99, p < 0.0001) and in strong concordance (limits of agreement from -1.9% to 1.0%, Bland-Altman analysis). In conclusion, a Vo2def coinciding with HRdef and VT was shown. This confirms that the determination of the WR-HR relationship and of HRdef is a practical and noninvasive means of identifying anaerobic threshold.     

Performance changes during a college playing career in NCAA division III football athletes

The purpose of this study was to compare anthropometric and athletic performance variables during the playing career of NCAA Division III college football players. Two hundred and eighty-nine college football players were assessed for height, body mass, body composition, 1-repetition-maximum (1RM) bench press, 1RM squat, vertical jump height (VJ), vertical jump peak, and vertical jump mean (VJMP) power, 40-yd sprint speed (40S), agility, and line drill (LD) over an 8-year period. All testing occurred at the beginning of summer training camp in each of the seasons studied. Data from all years of testing were combined. Players in their fourth and fifth (red-shirt year) seasons of competition were significantly (p < 0.05) heavier than first-year players. Significant increases in strength were seen during the course of the athletes' collegiate career (31.0% improvement in the 1RM bench press and 36.0% increase in squat strength). The VJ was significantly greater during the fourth year of competition compared to in the previous 3 years of play. Vertical jump peak and VJMP were significantly elevated from years 1 and 2 and were significantly higher during year 4 than during any previous season of competition. No significant changes in 40S or LD time were seen during the athletes playing career. Fatigue rate for the LD (fastest time/slowest time of 3 LD) significantly improved from the first (83.4 ± 6.4%) to second season (85.1 ± 6.5%) of competition. Fatigue rates in the fourth (88.3 ± 4.8%) and fifth (91.2 ± 5.2%) seasons were significantly greater than in any previous season. Strength and power performance improvements appear to occur throughout the football playing career of NCAA Division III athletes. However, the ability to significantly improve speed and agility may be limited.     

A retrospective analysis of American football hyperthermia deaths in the United States

Over the period 1980–2009, there were 58 documented hyperthermia deaths of American-style football players in the United States. This study examines the geography, timing, and meteorological conditions present during the onset of hyperthermia, using the most complete dataset available. Deaths are concentrated in the eastern quadrant of the United States and are most common during August. Over half the deaths occurred during morning practices when high humidity levels were common. The athletes were typically large (79% with a body mass index >30) and mostly (86%) played linemen positions. Meteorological conditions were atypically hot and humid by local standards on most days with fatalities. Further, all deaths occurred under conditions defined as high or extreme by the American College of Sports Medicine using the wet bulb globe temperature (WBGT), but under lower threat levels using the heat index (HI). Football-specific thresholds based on clothing (full football uniform, practice uniform, or shorts) were also examined. The thresholds matched well with data from athletes wearing practice uniforms but poorly for those in shorts only. Too few cases of athletes in full pads were available to draw any broad conclusions. We recommend that coaches carefully monitor players, particularly large linemen, early in the pre-season on days with wet bulb globe temperatures that are categorized as high or extreme. Also, as most of the deaths were among young athletes, longer acclimatization periods may be needed.     

The impact of different warm-up protocols on vertical jump performance in male collegiate athletes

The purpose of this study was to compare the impact of different types of warm-up on countermovement vertical jump (VJ) performance. Sixty-four male Division I collegiate football players completed a pretest for VJ height. The participants were then randomly assigned to a warm-up only condition, a warm-up plus static stretching condition, a warm-up plus dynamic stretching condition, or a warm-up plus dynamic flexibility condition. VJ performance was tested immediately after the completion of the warm-up. The results showed that there was a significant difference (P < .05) in VJ performance between the warm-up groups. Posttest jump performance improved in all groups; however, the mean for the static stretching group was significantly lower than the means for the other 3 groups. The static stretching negated the benefits gained from a general warm-up when performed immediately before a VJ test.     

Gender differences in hockey players during on-ice graded exercise

The purpose of this study was to examine whether gender differences exist for ventilatory threshold (VT), lactate threshold (LT), and Vo2max during on-ice skating in college hockey players. Ten male and 10 female Division III college hockey players performed a graded exercise skating protocol until reaching volitional fatigue. The graded exercise test employed stages that were 80 seconds in duration, with 40 seconds of rest between each stage to obtain blood lactate samples. Ventilatory threshold occurred at a higher percentage of maximal heart rate (HRmax) in women than in men. The women's VT occurred at 77.3% +/- 1.6% HRmax, while the men's VT occurred at 72.6% +/- 2.0% HRmax (p < 0.02). Men and women had similar HRmax values: 191.3 +/- 2.5 b.min and 185.8 +/- 2.5 b.min, respectively. Vo2max was different between genders, with men at 52.7 +/- 1.3 mL.kg.min and women at 40.1 +/- 1.0 mL.kg.min (p < 0.01). In addition, VT was different between genders when measured as a percentage of Vo2max, with men at 52.7% +/- 3.2% and women at 67.3% +/- 4.0% (p < 0.02). In contrast, LT was similar between genders when expressed as a percentage of HRmax or Vo2max. For each gender, LT occurred at a significantly higher percentage of HRmax or Vo2max than VT did. It can be concluded that VT does not accurately predict LT in male or female hockey players. Additionally, competitive female hockey players have a lower Vo2max but a higher VT than their male counterparts. An increased VT may be a compensatory mechanism to offset the smaller Vo2max values measured in female hockey players. On-ice testing is a practical way to address specific aerobic training needs of hockey players.     

Intensity-dependent tolerance to exercise after attaining V(O2) max in humans.

The tolerable duration of high-intensity, constant-load cycle ergometry is a hyperbolic function of power, with an asymptote termed critical power (CP) and a curvature constant (W') with units of work. It has been suggested that continued exercise after exhaustion may only be performed below CP, where predominantly aerobic energy transfer can occur and W' can be partially replenished. To test this hypothesis, six volunteers each performed cycle-ergometer exercise with breath-by-breath determination of ventilatory and pulmonary gas exchange variables. Initially, four exercise tests to exhaustion were made: 1). a ramp-incremental and 2). three high-intensity constant-load bouts at different work rates, to estimate lactate (theta(L)) and CP thresholds, W', and maximum oxygen uptake (Vo2 max). Subsequently, subjects cycled to the limit of tolerance (for approximately 360 s) on three occasions, each followed by a work rate reduction to 1). 110% CP, 2). 90% CP, and 3). 80% theta(L) for a 20-min target. W' averaged 20.9 +/- 2.35 kJ or 246 +/- 30 J/kg. After initial fatigue, 110% CP was tolerated for only 30 +/- 12 s. Each subject completed 20 min at 80% theta(L), but only two sustained 20 min at 90% CP; the remaining four subjects fatigued at 577 +/- 306 s, with oxygen consumption at 89 +/- 8% Vo2 max. The results support the suggestion that replenishing W' after fatigue necessitates a sub-CP work rate. The variation in subjects' responses during 90% CP was unexpected but consistent with mechanisms such as reduced CP consequent to prior high-intensity exercise, variation in lactate handling, and/or regional depletion of energy substrates, e.g., muscle glycogen.