We followed a cohort of 124 subjects with a history of inflammatory bowel disease to ascertain risk estimates for clinically active disease associated with exposure to recent stress events. We calculated risk estimates for three lag models (-1, 0, + 1 month). The data indicated a strong association between stress exposures and new clinical episodes of disease (RR = 2.9, 95% Cl: 2.0-4.1), most apparent in the immediate period (lag = 0). Risk estimates were also elevated for extended episodes of disease in subjects under stress compared with unexposed subjects. These results underscore the importance of monitoring stress exposures in prevention and treatment of recurrent disease.