Objective-To evaluate the effect of delayed exposure of dairy cattle to Mycobacterium avium subsp paratuberculosis (MAP) on the incidence of those cows testing positive for MAP and developing clinical Johne's disease (CJD). Animals-79 cows not exposed to MAP as calves (unexposed cohort) and 260 cows exposed to MAP as calves (exposed cohort). Procedures-Cows in the unexposed cohort were born into 5 MAP-uninfected herds and introduced at various ages into 5 MAP-infected herds where the exposed cohort cows were born and raised. Beginning when each cow was 24 months old, fecal and serum samples were collected annually from 2003 through 2006. Feces were cultured for MAP, and an ELISA was used to analyze serum samples for antibodies against MAP. Date and reason for culling were obtained from herd records. Incidence of positive culture and ELISA results and CJD was compared between unexposed and exposed cohort cows with Cox regression. Results-Compared with exposed cohort cows, the hazard ratios for unexposed cohort cows having positive culture results, having positive ELISA results, and developing CJD were 0.12, 0.03, and 0.001, respectively, and those ratios increased by 2%, 6%, and 17%, respectively, for each month spent in an MAP-infected herd. Conclusions and Clinical Relevance-Delayed exposure of cows to MAP resulted in lower incidences of positive culture and ELISA results and CJD in those cows, compared with incidences of cows exposed to MAP since birth. The hazard of testing positive for MAP or developing CJD increased with time, regardless of cohort.