The double-blind, prospective, National Surgical Adjuvant Breast and Bowel Project (NSABP) Breast Cancer Prevention Trial (BCPT) showed a 50% reduction in the risk of breast cancer for tamoxifen versus placebo, yet many women at risk of breast cancer do not adhere to the 5-year course. This first report of the rich BCPT drug adherence data examines predictors of adherence. Between June, 1992 and September, 1997 13,338 women at high risk of breast cancer were randomly assigned to 20 mg/d tamoxifen versus placebo; we analyzed the 11,064 enrolled more than 3 years before trial unblinding. Primary endpoint was full drug adherence (100% of assigned pills per staff report, excluding protocol-required discontinuation) at 1 and 36 months; secondary was adequate adherence (76%-100%). Protocol-specified multivariable logistic regression tested lifestyle factors, controlling for demographic and medical predictors. About 13% were current smokers; 60% were overweight/obese; 46% had moderate/heavy physical activity; 21%, 66%, 13% drank 0, 0-1, 1+ drinks per day, respectively; 91% were adequately adherent at 1 month; and 79% were at 3 years. Alcohol use was associated with reduced full adherence at 1 month (P = 0.016; OR = 0.79 1+ vs. 0), as was college education (P <0.001; OR = 0.78 vs. high school); age (P < 0.001; OR = 1.4 age 60+) and per capita household annual income (P < 0.001; OR = 1.2 per $30,000) with increased adherence. Current smoking (P = 0.003; OR = 0.75), age (P = 0.024, OR = 1.1), college education (P = 0.037; OR = 1.4), tamoxifen assignment (P = 0.031; OR = 0.84), and breast cancer risk (P <.001; OR = 1.5 high vs. low) predicted adequate adherence at 36 months. There were no significant associations with obesity or physical activity. Alcohol use and smoking might indicate a need for greater adherence support.