The zoonotic potential of Cryptosporidium was studied in one of the most densely populated provinces of Egypt regarding livestock and people. In a representative survey, faecal samples from cattle, buffalo and stool samples from diarrhoeic children (<10 years) were investigated. Parameters assumed to be related to cryptosporidiosis were recorded for animals and children. Animal samples (804) were examined by the Copro-antigen RIDA(®)QUICK test, followed by PCRs targeting the 18S rDNA and gp60 genes for antigen-positive and 10% randomly selected negative samples. All 165 human samples were tested by both methods. The overall estimated prevalence of Cryptosporidium in ruminants was 32.2%, without significant difference between animal species. PCR identified 65.7% Cryptosporidium parvum, 11.8% Cryptosporidium ryanae, 4.1% Cryptosporidium bovis, and combinations of C. parvum plus C. ryanae (11.2%), C. parvum plus C. bovis (5.3%) and of C. parvum plus Cryptosporidium andersoni (1.8%), also without significant differences in species occurrence between cattle and buffalos. The human Cryptosporidium spp. prevalence was 49.1%, of which 60.5% were Cryptosporidium hominis, 38.2% C. parvum and 1.2% C. parvum plus C. bovis. Analysis of gp60 variants allocated C. parvum found in animals to the zoonotic subtype family IIa (18.9%, subtype IIaA15G1R1 only) and to IId (81.1%, mostly IIdA20G1). In humans 50% were classified as subtype family IIa (IIaA15G1R1 and IIaA15G2R1) and 50% were IIdA20G1. C. andersoni occurred only in cattle older than 1 year. In contrast, mono-infections with one of the three single Cryptosporidium species and the three combinations with C. parvum were more prevalent in cattle and buffaloes younger than 1 year, particularly in those younger than 3 months, and were predominantly subtype family IId. In human samples no Cryptosporidium were identified in children younger than 7 months. Neither place of residence nor the source of drinking-water had measurable effects on prevalence. Remarkably, however, all children with C. parvum subtype family IIa and 86% with subtype family IId had contact to animals. High prevalence and identical genotypes of C. parvum in animals and humans indicate zoonotic transmission due to contact with animals, involving IIdA20G1 as the most frequent subtype.