A study was conducted to determine the effects of dietary factors in natural ingredient and purified diets on uterine weights of immature CD-1 mice used in uterotrophic bioassays. Factors evaluated included body weight gain, dietary phytoestrogen content, total metabolizable energy, and percent crude fiber. Fifteen to 147 mice per group, housed 5 per cage, were randomly assigned to each of the 20 test diets. The test diets were fed for 7 days to 15-day old immature female CD-1 mice and their body weight gain and uterine weights were determined. Analysis of covariance procedures were used to evaluate differences in uterine weights, after adjusting for body weight and time-related trends. Fisher's least significant difference test was used to compare adjusted uterine weights and weight gains among the test diets. Additionally, multiple linear regression procedures were used to identify those characteristics of the rodent diet that were most predictive of the adjusted uterine weights. Total metabolizable energy was significantly (P < 0.01) correlated with and was predictive of uterine weights. The following dietary variables were not significantly predictive of uterine weights: total daidzein and genistein content, percent protein, fat, N-FE (carbohydrates) or percent crude fiber. We concluded that: (1) total metabolizable energy (ME) in natural ingredient or purified diets has a significant (P < 0.01) effect on the uterine weights of immature mice used in 7-day uterotrophic bioassays; (2) a standardized, estrogen-free diet with a constant level of ME should be used for conducting uterotrophic assays when comparing results between different laboratories or when determining the estrogenic or anti-estrogenic activity of endocrine disruptor compounds; (3) the mouse uterotrophic assay remains a sensitive bioassay for assessing chemicals for estrogenic activity or for the detection of total estrogenic activity in rodent diets that may be contaminated with estrogenic compounds, and (4) chemical assays should be used to detect or measure low levels of the phytoestrogens in rodent diets.