Systems for exposing mice to 2,450-MHz electromagnetic fields

Two systems for exposing mice to 2,450-MHz electromagnetic fields are described. In a waveguide system, four mice were placed in a Styrofoam cage and exposed dorsally to circularly polarized electromagnetic fields. The temperature and humidity in the mouse holder were kept constant by forced-air ventilation. For 1-W input power to the waveguide, the average specific absorption rate (SAR) was determined by twin-well calorimetry to be 3.60 ± 0.11 (SE) W/kg in 27-g mice. The maximum SAR at the skin surface determined thermographically was 8.36 W/kg in the head of the mouse. The second system was a miniature anechoic chamber. Six mice were irradiated dorsally to far field plane waves. Copper shielding and high-temperature absorbing material were lined inside the chamber to accommodate the high input power. The air ventilation at the location of the mice was separately controlled so that any heating in the absorber would not affect the animals. For 1-W input power, the average SAR was 0.17 ± 0.01 W/kg and the maximum SAR at the skin surface was 0.41 W/kg in the animal when irradiated with body axis parallel to the E field; the SARs were 0.11 ± 0.01 W/kg and 0.64 W/kg, respectively, when irradiated perpendicular to the E field.     

Estimation of whole‐body SAR from electromagnetic fields using personal exposure meters

In this article, personal electromagnetic field measurements are converted into whole‐body specific absorption rates for exposure of the general public. Whole‐body SAR values calculated from personal exposure meter data are compared for different human spheroid phantoms: the highest SAR values (at 950 MHz) are obtained for the 1‐year‐old child (99th percentile of 17.9 µW/kg for electric field strength of 0.36 V/m), followed by the 5‐year‐old child, 10‐year‐old child, average woman, and average man. For the 1‐year‐old child, whole‐body SAR values due to 9 different radiofrequency sources (FM, DAB, TETRA, TV, GSM900 DL, GSM1800 DL, DECT, UMTS DL, WiFi) are determined for 15 different scenarios. An SAR matrix for 15 different exposure scenarios and 9 sources is provided with the personal field exposure matrix. Highest 95th percentiles of the whole‐body SAR are equal to 7.9 µW/kg (0.36 V/m, GSM900 DL), 5.8 µW/kg (0.26 V/m, DAB/TV), and 7.1 µW/kg (0.41 V/m, DECT) for the 1‐year‐old child, with a maximal total whole‐body SAR of 11.5 µW/kg (0.48 V/m) due to all 9 sources. All values are below the basic restriction of 0.08 W/kg for the general public. 95th percentiles of whole‐body SAR per V/m are equal to 60.1, 87.9, and 42.7 µW/kg for GSM900, DAB/TV, and DECT sources, respectively. Functions of the SAR versus measured electric fields are provided for the different phantoms and frequencies, enabling epidemiological and dosimetric studies to make an analysis in combination with both electric field and actual whole‐body SAR. Bioelectromagnetics 31:286–295, 2010. © 2009 Wiley‐Liss, Inc.     

Dosimetry evaluation of a cylindrical waveguide chamber for unrestrained small rodents at 1.9 GHz

An exposure system, consisting of four identical cylindrical waveguide chambers, was developed for studying the effects of radiofrequency (RF) energy on laboratory mice at a frequency of 1.9 GHz. The chamber was characterized for RF dose rate as a function of animal body mass and dose rate variations due to animal movement in the cage. Dose rates were quantified in terms of whole‐body average (WBA) specific absorption rate (SAR), brain average (BA) SAR and peak spatial‐average (PSA) SAR using measurement and computational methods. Measurements were conducted on mouse cadavers in a multitude of possible postures and positions to evaluate the variations of WBA‐SAR and its upper and lower bounds, while computations utilizing the finite‐difference time‐domain method together with a heterogeneous mouse model were performed to determine variations in BA‐SAR and the ratio of PSA‐SAR to WBA‐SAR. Measured WBA‐SAR variations were found to be within the ranges of 9–23.5 W/kg and 5.2–13.8 W/kg per 1 W incident power for 20 and 40 g mice, respectively. Computed BA‐SAR variations were within the ranges of 3.2–10.1 W/kg and 3.3–9.2 W/kg per 1 W incident power for 25 and 30 g mouse models, respectively. Ratios of PSA‐SAR to WBA‐SAR, averaged over 0.5 mg and 5 mg tissue volumes, were observed to be within the ranges of 6–15 and 4–10, respectively. Bioelectromagnetics 33:575–584, 2012. © 2012 Wiley Periodicals, Inc.     

Dosimetry in mice exposed to 1.6 GHz microwaves in a carrousel irradiator

We have developed a carrousel irradiator for mice which delivers a head‐first and near‐field radiofrequency exposure that more closely simulates cellular telephone and radio use than conventional whole body exposure systems. Mouse cadavers were placed on the carrousel irradiator and exposed with their noses 5 mm from the feedpoint of a 1.6 GHz antenna. Local measured specific absorption rates (SAR) in brain regions corresponding to the frontal cortex, medial caudate putamen, and midhippocampal areas were 2.9, 2.4, and 2.2 W/kg per watt of irradiated power, respectively. In addition, average SAR was estimated to be 3.4 W/kg per watt along the sagittal plane of the brain, 2.0 W/kg per watt along the sagittal plane of the body, and between 6.8 and 8.1 W/kg per watt at peak locations along the sagittal plane at the body surface. This detailed SAR information in mice is critical to the interpretation of biological studies of IRIDIUM exposure, and similar analysis should be included for all studies of in vivo exposure of small animals to microwaves. Bioelectromagnetics 20:42–47, 1999. © 1999 Wiley‐Liss, Inc.     

Millimeter wave power density in aqueous biological samples

Power density distribution inside a water sample placed between two parallel lossy dielectric plates (Polystyrene) was calculated using Fresnel equations for the frequency range of 42.25-53.57 GHz. Due to the multiple internal reflections from the sample boundaries, the distribution of the power density within the thin sample is more uniform than that within a semi-infinite medium. The power density in a sample depends on the thicknesses of the sample and the adjacent dielectric plates. For the given frequency range the sample thickness optimal for power density uniformity varies between 0.28 and 0.33 mm. The front plate has a significant effect on the magnitude of the power density within the sample but little effect on the power density distribution. The thicker the rear plate, the greater is the non uniformity of the power density distribution within the sample. Based on the calculated data, we determined the dimension of an exposure chamber providing the optimal power density distribution uniformity for mm-wave irradiation.     

Effects of gestational exposure to 1.95‐GHz W‐CDMA signals for IMT‐2000 cellular phones: Lack of embryotoxicity and teratogenicity in rats

The present study was designed to evaluate whether gestational exposure to an EMF targeting the head region, similar to that from cellular phones, might affect embryogenesis in rats. A 1.95‐GHz wide‐band code division multiple access (W‐CDMA) signal, which is one applied for the International Mobile Telecommunication 2000 (IMT‐2000) system and used for the freedom of mobile multimedia access (FOMA), was employed for exposure to the heads of four groups of pregnant CD(SD) IGS rats (20 per group) for gestational days 7–17. The exposure was performed for 90 min/day in the morning. The spatial average specific absorption rate (SAR) for individual brains was designed to be 0.67 and 2.0 W/kg with peak brain SARs of 3.1 and 7.0 W/kg for low (group 3) and high (group 4) exposures, respectively, and a whole‐body average SAR less than 0.4 W/kg so as not to cause thermal effects due to temperature elevation. Control and sham exposure groups were also included. At gestational day 20, all dams were killed and fetuses were taken out by cesarean section. There were no differences in maternal body weight gain. No adverse effects of EMF exposure were observed on any reproductive and embryotoxic parameters such as number of live (243–271 fetuses), dead or resorbed embryos, placental weights, sex ratios, weights or external, visceral or skeletal abnormalities of live fetuses. Bioelectromagnetics 30:205–212, 2009. © 2008 Wiley‐Liss, Inc.     

Comparison of dose dependences for bioeffects of continuous-wave and high-peak power microwave emissions using gel-suspended cell cultures

The study compared bioeffects of continuous wave (CW) microwaves and short, extremely high power pulses (EHPP) at the same carrier frequency (9.3 GHz) and average power (1.25 W). The peak transmitted power for EHPP was 250 kW (0.5-micro s pulse width, 10 p.p.s.), producing the E field of 1.57 MV/m in the waveguide. A biological endpoint was the density of yeast cells, achieved after a 6 h growth period in a solid nutrient medium (agarose gel) during EHPP or CW exposure. Owing to power losses in the medium, the specific absorption rate (SAR) ranged from 3.2 kW/kg at the exposed surface of the sample to 0.6 mW/kg at 24 mm depth. Absorption and penetration of EHPP was identical to CW, producing peak SAR values 200 000 times higher than the average SAR, as high as 650 MW/kg at the surface. CW and EHPP exposures produced highly nonuniform but identical heating patterns in exposed samples. Following the exposure, the samples were sliced in a plane perpendicular to the wave propagation, in order to separate cell masses exposed at different SAR levels. Cell density in the slices was determined by nephelometry and compared to unexposed parallel control samples. Cell density was strongly affected by irradiation, and the changes correlated well with the local temperature rise. However, the data revealed no statistically significant difference between CW and EHPP samples across the entire studied range of SAR levels (over six orders of magnitude). A trend (P<0.1) for such a difference was observed in slices that were exposed at a time average SAR of 100 W/kg and higher, which corresponded to peak SAR above 20 MW/kg for the EHPP condition. These numbers could be indicative of a threshold for a specific (not merely thermal) exposure effect if the trend is confirmed by future studies.     

Reevaluation and improved design of the TEM cell in vitro exposure unit for replication studies

The transverse electromagnetic (TEM) cell system developed by Litovitz et al. and utilized by Penafiel et al. for the exposure of cells in T25 flasks at 835 MHz has been reevaluated for the purpose of replicating the studies published by Penafiel. The original setup has been reconstructed as closely as possible, with improvements enabling blinded exposures, forced cooling and better repeatable positioning of the flasks, as well as tight exposure and environmental parameter control. The signal unit can simulate the original signal but also enables various other exposure schemes. The setup has been evaluated for four T25 flasks filled with 5 and 10 ml of cell medium by experimental and numerical means. Comparing E field, SAR and temperature measurements resulted in good agreement: <0.4 dB (4.5%) for E field and 0.48 dB (10.5%) for SAR. The overall average SAR within the medium is 6.0 W/kg at 1 W input power with a standard deviation of less than 52%. The temperature increase was determined to be 0.13 degrees C/(W/kg). This can be reduced to 0.045 degrees C/(W/kg) by applying active air flow cooling. The comparison of SAR values from temperature measurements with the corresponding simulated values resulted in excellent agreement. These results do not correspond to the previous study reporting an average SAR within the medium of 2.5 W/kg at an input power of 0.96 W.     

Effect of electromagnetic radiofrequency radiation on the rats' brain, liver and kidney cells measured by comet assay

The goal of study was to evaluate DNA damage in rat's renal, liver and brain cells after in vivo exposure to radiofrequency/microwave (Rf/Mw) radiation of cellular phone frequencies range. To determine DNA damage, a single cell gel electrophoresis/comet assay was used. Wistar rats (male, 12 week old, approximate body weight 350 g) (N = 9) were exposed to the carrier frequency of 915 MHz with Global System Mobile signal modulation (GSM), power density of 2.4 W/m2, whole body average specific absorption rate SAR of 0.6 W/kg. The animals were irradiated for one hour/day, seven days/week during two weeks period. The exposure set-up was Gigahertz Transversal Electromagnetic Mode Cell (GTEM--cell). Sham irradiated controls (N = 9) were apart of the study. The body temperature was measured before and after exposure. There were no differences in temperature in between control and treated animals. Comet assay parameters such as the tail length and tail intensity were evaluated. In comparison with tail length in controls (13.5 +/- 0.7 microm), the tail was slightly elongated in brain cells of irradiated animals (14.0 +/- 0.3 microm). The tail length obtained for liver (14.5 +/- 0.3 microm) and kidney (13.9 +/- 0.5 microm) homogenates notably differs in comparison with matched sham controls (13.6 +/- 0.3 microm) and (12.9 +/- 0.9 microm). Differences in tail intensity between control and exposed animals were not significant. The results of this study suggest that, under the experimental conditions applied, repeated 915 MHz irradiation could be a cause of DNA breaks in renal and liver cells, but not affect the cell genome at the higher extent compared to the basal damage.     

1439 MHz pulsed TDMA fields affect performance of rats in a T-maze task only when body temperature is elevated

This study sought to clarify the effects of exposure to electromagnetic waves (EMW) used in cellular phones on learning and memory processes. Sprague-Dawley rats were exposed for either 1 h daily for 4 days or for 4 weeks to a pulsed 1439 MHz time division multiple access (TDMA) field in a carousel type exposure system. At the brain, average specific absorption rate (SAR) was 7.5 W/kg, and the whole body average SAR was 1.7 W/kg. Other subjects were exposed at the brain average SAR of 25 W/kg and the whole body average SAR of 5.7 W/kg for 45 min daily for 4 days. Learning and memory were evaluated by reversal learning in a food rewarded T-maze, in which rats learned the location of food (right or left) by using environmental cues. The animals exposed to EMW with the brain average SAR of 25 W/kg for 4 days showed statistically significant decreases in the transition in number of correct choices in the reversal task, compared to sham exposed or cage control animals. However, rats exposed to the brain average SAR of 7.5 W/kg for either 4 days or for 4 weeks showed no T-maze performance impairments. Intraperitoneal temperatures, as measured by a fiber optic thermometer, increased in the rats exposed to the brain average SAR of 25 W/kg but remained the same for the brain average SAR of 7.5 W/kg. The SAR of a standard cellular phone is restricted to a maximum of 2 W/kg averaged over 10 g tissue. These results suggest that the exposure to a TDMA field at levels about four times stronger than emitted by cellular phones does not affect the learning and memory processes when there are no thermal effects.     

Acute effects of pulsed microwaves and 3-nitropropionic acid on neuronal ultrastructure in the rat caudate-putamen

Ultrastructure of the medium sized "spiny" neuron in rat dorsal-lateral caudate-putamen was assessed after administration of 3-nitropropionic acid (3-NP) and exposure to pulsed microwaves. Sprague-Dawley male rats were given two daily intraperitoneal doses of 0 or 10 mg/kg 3-NP and 1.5 h after each dose were exposed to microwave radiation at a whole body averaged specific absorption rate (SAR) of 0 (sham exposure), 0.6, or 6 W/kg for 30 min. Microwave exposure consisted of 1.25 GHz radiation delivered as 5.9 micros pulses with repetition frequency 10 Hz. Tissue samples taken 2-3 h after the second sham or microwave exposure showed no injury with light microscope methods. Blinded qualitative assessment of ultrastructure of randomly selected neurons from the same samples did reveal differences. Subsequent detailed, quantitative measurements showed that, when followed by sham exposure, administration of 3-NP significantly increased endoplasmic reticulum (ER) intracisternal width, ER area density, and nuclear envelope thickness. Microwave exposure at 6 W/kg alone also significantly increased these measures. Exposure of 3-NP treated animals at 6 W/kg significantly increased effects of 3-NP on ultrastructure. Although exposure at 0.6 W/kg alone did not affect ultrastructure measures, exposure of 3-NP treated animals at 0.6 W/kg reduced the effects of 3-NP. We concluded that 3-NP changed neuronal ultrastructure and that the microwave exposures used here changed neuronal ultrastructure in ways that depended on microwave SAR and neuron metabolic status. The apparent cancellation of 3-NP induced changes by exposure to pulsed microwaves at 0.6 W/kg indicated the possibility that such exposure can protect against the effects of mitochondrial toxins on the nervous system.     

Acute exposure to pulsed 2450-MHz microwaves affects water-maze performance of rats

Rats were trained in six sessions to locate a submerged platform in a circular water maze. They were exposed to pulsed 2450-MHz microwaves (pulse width 2 micros, 500 128;pulses/s, average power density 2 mW/cm(2), average whole body specific absorption rate 1.2 W/kg) for 1 h in a circular waveguide system immediately before each training session. One hour after the last training session, they were tested in a probe trial during which the platform was removed and the time spent in the quadrant of the maze in which the platform had been located during the 1-min trial was scored. Three groups of animals, microwave-exposed, sham-exposed, and cage control, were studied. Microwave-exposed rats were slower than sham-exposed and cage control rats in learning to locate the platform. However, there was no significant difference in swim speed among the three groups of animals, indicating that the difference in learning was not due to a change in motor functions or motivation. During the probe trial, microwave-exposed animals spent significantly less time in the quadrant that had contained the platform, and their swim patterns were different from those of the sham-exposed and cage control animals. The latter observation indicates that microwave-exposed rats used a different strategy in learning the location of the platform. These results show that acute exposure to pulsed microwaves caused a deficit in spatial "reference" memory in the rat.     

Intermittent exposure of rats to 2450 MHz microwaves at 2.5 mW cm2: behavioral and physiological effects

Long-Evans male adult rats were intermittently exposed for 14 weeks to continuous wave (CW) 2450-MHz microwaves at an average power density of 2.5 mW/cm2. The mean specific absorption rate was 0.70 W/kg (+/- 0.02 SEM). The rats were exposed 7 h/day, 7 days/week in a radiation chamber with a monopole above ground, while housed in Plexiglas cages. Weekly measures of body mass and food intake did not indicate statistically significant effects of microwave irradiation. Assessments of threshold for electric-footshock detection revealed a significant difference between microwave and sham-exposed animals. Assessments of cholinesterase and sulfhydryl groups in blood and 17-ketosteroids in urine did not distinguish the two groups of rats. Behavioral measures made at the end of the 14-week exposure included an open-field test, shuttlebox avoidance performance, and schedule-controlled lever-pressing for food pellets. Statistically significant differences between microwave- and sham-exposed rats were observed for these measures. Examination of adrenal tissue, plasma electrolytes, and organ masses after 14 weeks of exposure revealed no difference between the two groups of rats.     

No evidence of major transcriptional changes in the brain of mice exposed to 1800 MHz GSM signal

To analyze possible effects of microwaves on gene expression, mice were exposed to global system for mobile communication (GSM) 1800 MHz signal for 1 h at a whole body SAR of 1.1 W/kg. Gene expression was studied in the whole brain, where the average SAR was 0.2 W/kg, by expression microarrays containing over 22,600 probe sets. Comparison of data from sham and exposed animals showed no significant difference in gene expression modulation. However, when less stringent constraints were adopted to analyze microarray results, 75 genes were found to be modulated following exposure. Forty-two probes showed fold changes ranging from 1.5 to 2.8, whereas 33 were down-regulated from 0.67- to 0.29-fold changes, but these differences in gene expression were not confirmed by real-time PCR. Under these specific limited conditions, no consistent indication of gene expression modulation in whole mouse brain was found associated to GSM 1800 MHz exposure.     

Normalizing the thermal effects of radiofrequency radiation: body mass versus total body surface area

The current guideline for exposure to radiofrequency radiation (RFR) was developed through assessment of the biological effects data collected primarily from the rat. The consensus that a lack of hazardous biological effects occurred below a whole-body-averaged specific absorption rate (SAR) of 4.0 W/kg led to the proposition of a 0.4 W/kg guideline with a built-in safety factor of 10. This paper demonstrates that if the RFR absorption rate in the rat had been normalized with respect to total body surface area rather than body mass, the exposure guideline would be 2.3 W/m2, which translates to an SAR of approximately 0.06 W/kg for an adult human. It is further shown that a given RFR absorption rate, normalized as a fraction of a species' heat loss per unit of surface area, is independent of body mass over a range of 0.03-100 kg; however, a normalization of the RFR absorption rate to heat loss per unit of body mass is highly dependent on the species' mass. Normalizing the rate of RFR absorption to the surface area of the rat indicates that the current RFR exposure guideline of 0.4 W/kg may be too high.     

Behavioral and physiological effects of chronic 2,450-MHz microwave irradiation of the rat at 0.5 mW/cm2

Adult male Long-Evans rats were intermittently exposed to 2450 MHz CW microwaves at an average power density of 0.5 mW/cm2 for 90 days. The resulting SAR was 0.14 W/kg (range 0.11 to 0.18 W/kg). The animals were exposed 7 h/day, 7 days/wk, for a total of 630 h in a monopole-above-ground radiation chamber while housed in Plexiglas holding cages. Daily measures of body mass and food and water intake indicated no statistically significant effects of microwave exposure. Monthly assessment of reactivity to electric footshock, levels of cholinesterase and sulfhydryl groups in blood, and 17-ketosteroids in urine revealed no reliable differences between 14 sham-exposed and 14 microwave-exposed rats. After the 90 days of exposure, seven rats, randomly chosen from each group, were assessed for open-field behavior, shuttlebox performance, and schedule-controlled (IRT schedule) lever pressing for food pellets. Statistically significant differences between microwave-exposed and sham-exposed rats were observed in shuttlebox performances and lever pressing. Post mortem measures of mass of several organs and microscopic examination of adrenal tissue revealed no differences between the two groups of animals.     

Chronic exposure of rabbits to 0.5 and 5 mW/cm2 2450-MHz CW Microwave Radiation

Two groups of 16 male New Zealand rabbits were exposed to 2450-MHz continuous wave microwave fields in two experiments of 90 days each. The incident power densities of the first and second experiment were 0.5 and 5 mW/cm², respectively. During each study, 16 animals were adapted to a miniature anechoic chamber exposure system for at least 2 weeks, then 8 of them were exposed for 7 h daily, 5 days a week for 13 weeks, and the other 8 animals were sham exposed. The rabbits were placed in acrylic cages, and each was exposed from the top in an individual miniature anechoic chamber. Thermography showed a maximum specific absorption rate of 5.5 W/kg in the head and 7 W/kg in the back at 5-mW/cm² incident power density. After each 7-h exposure session, the animals were returned to their home cages. Food consumption in the exposure chamber and body mass were measured daily. Blood samples were taken before exposure and monthly thereafter for hematological, morphological, chemical, protein electrophoresis, and lymphocyte blast transformation studies. Eyes were examined for cataract formation. Finally, pathological examinations of 28 specimens of organs and tissues of each rabbit were performed. Statistically, there was a significant (P < .01) decrease only of food consumption during the 5-mW/cm² exposure; other variables were not significantly different between exposed and control groups.     

Specific absorption rate in rats exposed to 2,450-MHz microwaves under seven exposure conditions

Both positive and negative biological effects of microwaves on drug actions in rats exposed to 1-mW/cm2, 2,450-MHz microwaves have been reported by several investigators. We conducted dosimetry studies for seven different exposure conditions to determine whether these different results could be due to the rats having been exposed differently. They included anterior and posterior exposures in a circular waveguide, near field, far field with E- or H-field parallel to the long axis of the body and dorsal exposure in a miniature anechoic chamber with E- or H-field parallel to the long axis of the body. The average specific absorption rates (SARs) in the head, tail, and body of the exposed rats were measured by means of a calorimetry system. The local SARs at eight locations in the brain were determined by temperature measurement with Vitek probes. Intensive coupling of energy to the tail when it was exposed parallel to the E-field was shown by thermography. For the same average incident power density, the average SARs in the heads of rats were about two times higher in the circular waveguide than for other exposures. The local SARs in the brain varied for different exposure conditions. Statistical comparisons of SARs under the different exposure conditions are presented.     

The effects of simultaneous combined exposure to CDMA and WCDMA electromagnetic fields on rat testicular function

Wireless mobile phones and other telecommunication devices are used extensively in daily life. We therefore examined the effects of combined exposure to radiofrequency electromagnetic fields (RF‐EMF) on rat testicular function, specifically with respect to sensitive processes such as spermatogenesis. Male rats were exposed to single code division multiple access (CDMA) and wideband code division multiple access (WCDMA) RF signals for 12 weeks. The RF exposure schedule comprised 45 min/day, 5 days/week for a total of 12 weeks. The whole‐body average specific absorption rate (SAR) of CDMA and WCDMA was 2.0 W/kg each or 4.0 W/kg in total. We then investigated the correlates of testicular function such as sperm count in the cauda epididymis, testosterone concentration in the blood serum, malondialdehyde concentrations in the testes and epididymis, frequency of spermatogenesis stages, and appearance of apoptotic cells in the testes. We also immunoblotted for p53, bcl2, GADD45, cyclin G, and HSP70 in the testes of sham‐ and combined RF‐exposed animals. Based on the results, we concluded that simultaneous exposure to CDMA and WCDMA RF‐EMFs at 4.0 W/kg SAR did not have any observable adverse effects on rat spermatogenesis. Bioelectromagnetics 33:356–364, 2012. © 2011 Wiley Periodicals, Inc.     

Effects of 2.45 GHz microwaves on meiotic chromosomes of male CBA/CAY mice

Male CBA/CAY mice were exposed daily (6 days a week) for 30 minutes in an environmentally controlled waveguide to continuous 2.45 GHz microwave radiation for 2 weeks at average whole body absorbed dose rates of 0.05, 0.5, 10, and 20 mW/g. Shan exposed animals served as controls. Chain translocations were observed at diakinesis at metaphase I in microwave exposed animals. The yield of translocations increased with exposure, and varied nonlinearly with dose rate. An increase in incidence of univalents was seen after exposure at 10 and 20 mW/g. The findings are interpreted to indicate interference with normal spermatogenesis during the exposure period.