Sasamoto N, et al. - Among premenopausal women (n=815) from the New England Case Control Study (NEC), researchers assessed predictors of cancer antigen 125 (CA125) levels by developing linear and dichotomous (≥35 U/mL) CA125 prediction models. An abridged model restricting to available predictors in 473 premenopausal women in the European Prospective Investigation into Cancer and Nutrition Study (EPIC) was externally validated. Age, race, tubal ligation, endometriosis, menstrual phase at blood draw, and fibroids were included in the final linear CA125 prediction model, which explained 7% of the total variance of CA125. They reported similar correlation coefficients for the correlation between observed and predicted CA125 levels based on the abridged model (comprising age, race, and menstrual phase at blood draw) in NEC (r = 0.22) and in EPIC (r = 0.22). In the detection of healthy women likely to have elevated CA125, the proposed model could have utility. This model can improve the specificity of CA125 for ovarian cancer screening.