Dr James' summaries of occupational associations with lowered sex ratios and the health effects of nonionizing radiation, and the hypothesis that change in a ratio of parental hormones will change the probability of a child's gender, are of interest. Our study could not address these questions inasmuch as it examined only potential hormonal and semen quality effects from specific occupational exposure to 3- to 100-MHz radio-frequency radiation. Our findings may not be generalizable to the effects of nonionizing radiation exposure at other frequencies. We agree with Dr James' recommendation that further work to determine the effects of nonionizing radiation on reproductive endpoints should take place in occupationally exposed groups. Successful additional research in this area is likely to be limited to study designs in occupational groups that possess adequate statistical power, definitive exposure metrics, adequate exposure assessment, control or adequate measurement of potential confounders and effect modifiers, and selection of sensitive, interpretable, and standardized reproductive outcomes. Studies based on very rare outcomes or very common outcomes (eg, altered sex ratio) require very large numbers of study participants to attain adequate statistical power, especially when control for potential confounders and effect modifiers is considered. This is difficult in occupational groups, which are often limited in size. The most effective measures of reproductive health for these studies may be sperm concentration, motility, and structure, hormonal measures of the reproductive profile in men, and hormonal measures of ovulatory function in women.