Several studies have reported an association between caffeine intake and delay to conception. To study this relation further, the authors examined caffeine use in 1,050 women with primary infertility and 3,833 women who had recently given birth during the period 1981-1983 in the United States and Canada. The cases were separated by the cause of their infertility: ovulatory factor, tubal disease, cervical factor, endometriosis, or idiopathic infertility. The relative risks of each type of infertility associated with caffeine were calculated using separate logistic regression models and controlling for relevant confounding factors, such as age, center, cigarette smoking, lifetime number of sexual partners, alcohol consumption, contraception, body mass index, and exercise. A significant increase in the risk of infertility due to tubal disease or endometriosis was observed for the upper levels of caffeine intake, indicating a threshold effect. For tubal infertility, a relative risk of 1.5 (95% confidence interval (CI) 1.1-2.0) was found in women who consumed more than 7 g of caffeine per month as compared with those who consumed 3 g or less per month. For endometriosis, the relative risk was 1.9 (95% CI 1.2-2.9) in women who consumed 5.1-7 g/month and 1.6 (95% CI 1.1-2.4) in those with an intake of more than 7 g/month. These data suggest that caffeine deserves further study with regard to its effects on the female reproductive system.