Effective management of human cryptosporidiosis requires efficient methods for detection and identification of the species of Cryptosporidium isolates. Identification of isolates to the species level is not routine for diagnostic assessment of cryptosporidiosis, which leads to uncertainty about the epidemiology of the Cryptosporidium species that cause human disease. We developed a rapid and reliable method for species identification of Cryptosporidium oocysts from human fecal samples using terminal restriction fragment polymorphism (T-RFLP) analysis of the 18S rRNA gene. This method generated diagnostic fragments unique to the species of interest. A panel of previously identified isolates of species was blind tested to validate the method, which determined the correct species identity in every case. The T-RFLP profiles obtained for samples spiked with known amounts of Cryptosporidium hominis and Cryptosporidium parvum oocysts generated the two expected diagnostic peaks. The detection limit for an individual species was 1% of the total DNA. This is the first application of T-RFLP to protozoa, and the method which we developed is a rapid, repeatable, and cost-effective method for species identification.