Similarities between Johne's disease in ruminants and Crohn's disease in humans have led to speculation that Mycobacterium avium paratuberculosis (MAP) might be a causative agent in Crohn's disease. However, evidence remains inconsistent. In this case-control study (1999-2004), the authors assessed the possible role of drinking water and dairy products potentially contaminated with MAP in the etiology of Crohn's disease. A total of 218 patients with Crohn's disease recruited from nine hospitals in England and 812 controls recruited from the community completed a short questionnaire for evaluation of proxy measures of potential exposure to MAP. Logistic regression showed no significant association with measures of potential contamination of water sources with MAP, water intake, or water treatment. Multivariate analysis showed that consumption of pasteurized milk (per kg/month: odds ratio (OR) = 0.82, 95% confidence interval (CI): 0.69, 0.97) was associated with a reduced risk of Crohn's disease. Meat intake (per kg/month: OR = 1.40, 95% CI: 1.17, 1.67) was associated with a significantly increased risk of Crohn's disease, whereas fruit consumption (per kg/month: OR = 0.78, 95% CI: 0.67, 0.92) was associated with reduced risk. This study does not support a role for water or dairy products potentially contaminated with MAP in the etiology of Crohn's disease. The observed association with meat and the negative association with pasteurized milk need further study.