Mouse sperm collected from the cauda epididymidis 35 days after acute testicular X-ray exposure and fluorescently stained for DNA show dose-dependent increases in the coefficient of variation (CV) of flow cytometrically obtained fluorescence distributions. By comparing dose-response curves obtained with three protocols which overcome the optical and cytochemical difficulties of sperm measurement in different ways we conclude the response is due to X-ray-induced DNA content variability. In the range between 0 and 600 rad the dose dependence of the square of CV of the DNA content variability, delta CV2D, is described by delta CV2D = Bx + Cx2, with 0 less than or equal to B less than or equal to 0.23 X 10(-2) and C = (0.44 +/- 0.06) X 10(-4). The dose x is measured in rad and delta CVD is expressed in percent. Computer modeling of the shapes of the fluorescence distributions show that at 600 rad 30 to 40% of the sperm have abnormal DNA content. Some have errors as large as two whole chromosomes, but it is not clear whether they are due to whole chromosome nondisjunction or a finer fragmentation of the genome. Exposures to benzo(a)pyrene and mitomycin C cause no detectable DNA content variability. We conclude mouse sperm DNA content measurements are not sensitive to small amounts of aneuploidy and as such will only be useful in detecting agents that produce substantial DNA content variability. Another animal with a smaller number of chromosomes might be more favorable. These sperm measurement techniques may find additional application in other areas of reproductive biology, such as the determination of the relative numbers of X and Y chromosome-bearing sperm in semen that may be artificially enriched in one population.