Data from the Ontario Farm Family Health Study were analyzed to determine whether smoking, caffeine, or alcohol use among men and women affect fecundability (the monthly probability of conception). In this retrospective cohort study of farm couples in Ontario, Canada, the farm operator, husband, and wife completed questionnaires during 1991-1992, yielding information on 2,607 planned pregnancies that had occurred over the previous 30 years. Fecundability ratios were calculated using an analog of the Cox proportional hazards model. Cigarette smoking among women and men was associated with decreased fecundability (fecundability ratio = 0.90, 95% confidence interval (CI) 0.82-0.98 and fecundability ratio = 0.88, 95% CI 0.81-0.95, respectively). Caffeine consumption of 100 mg or less versus more than 100 mg in women and men was not associated with fecundability (fecundability ratio = 0.98, 95% CI 0.91-1.07 and fecundability ratio = 1.05, 95% CI 0.97-1.14, respectively). Decreases were observed among women who were coffee drinkers (fecundability ratio = 0.92, 95% CI 0.84-1.00) and men who were heavy tea drinkers (fecundability ratio = 0.85, 95% CI 0.69-1.05), regardless of caffeine content. Alcohol use among women and men was not associated with fecundability. These data are consistent with previous studies of the adverse effect of tobacco on fecundability in female smokers and suggest an effect of smoking among males. Continued evaluation of coffee and tea is warranted to address constituents other than caffeine.