Abstract A longitudinal study of 2-year duration was conducted to determine the risk, as measured by incidence rate, of Cryptosporidium parvum infection among dairy cattle in the Catskill/Delaware Watershed of New York City (NYC), and the factors that predispose animals to the likelihood of infection. A proportional sampling scheme with follow up at quarterly farm visits was employed for heifers and cows. Additionally, all calves born on the 39 study farms were sampled once during the first four weeks of life and at least once more before weaning. Samples were analyzed for the presence of C. parvum using a quantitative centrifugation concentration flotation technique and a C. parvum-specific enzyme-linked immunosorbent assay (ELISA). Of the 9914 fecal samples collected, 747 were found to contain C. parvum. The average number of oocysts detected was 1.3 × 10 5/g (range: 1.0/g–8.2 × 10 6/g). The average age at time of first detection of the organism was 15.0 days with a standard deviation of 6.59 days. The age range of animals infected with C. parvum in the study population was 3–60 days (inclusive). The unadjusted (crude) incidence rate of C. parvum among the entire study population was 2.05 per 1000 animal-days. The unadjusted incidence rate among pre-weaned calves was 15.55 per 1000 animal-days. After controlling for age and prior protozoal risk level, no seasonal impact on the incidence of C. parvum was detected among animals less than 61 days by negative binomial regression. A seasonal impact was identified among the oocyst counts of infected animals after controlling for age and prior protozoal risk level.