OBJECTIVE To isolate and speciate Cryptosporidium DNA from fecal samples obtained from dairy cattle in New York State and identify factors associated with whether cattle were shedding Cryptosporidium parvum versus Cryptosporidium bovis. METHODS Cross-sectional study. METHODS 115 fecal samples positive for DNA coding for the Cryptosporidium 18S rRNA gene from dairy cattle in New York State. METHODS A PCR assay was used to amplify DNA from fecal samples; amplification products were submitted for bidirectional DNA sequencing. Logistic regression was used to test for associations between various host factors and Cryptosporidium spp. RESULTS 70 of the 115 (61%) fecal samples were found to have C parvum DNA, 42 (37%) were determined to have C bovis DNA, and 3 (3%) were found to have C parvum deer-type DNA. The presence of diarrhea at the time of fecal sample collection, oocyst count, and breed were associated with whether cattle were infected with C parvum or C bovis, with animals more likely to be infected with C parvum if they had diarrhea, had a high oocyst count, or were Holsteins. CONCLUSIONS Results suggest that C parvum and C bovis can be isolated from dairy cattle in New York State and that various factors affect whether cattle infected with Cryptosporidium spp are infected with C parvum or C bovis. Findings also lend credence to the theory that C bovis may be more host adapted and thus less pathogenic to dairy cattle than C parvum.