Five Cryptosporidium-free Canada geese (Branta canadensis) were individually orally dosed with 3.5 x 10(6) Cryptosporidium parvum oocysts infectious to neonatal BALB/c mice. After intestinal passage, inoculum-derived oocysts extracted from goose faeces established severe infection in 14 neonatal BALB/c mice (inoculum dose 2.5 x 10(5)/mouse). The inoculum-derived oocysts were detected in goose faeces up to 9 days post-inoculation (PI); the number of intact oocysts and oocyst shells shed during the first 3 days PI was significantly higher than for the remaining 6 days PI (P < 0.01). Based on acid-fast stained air-dried direct wet smears, 62% of the oocysts in goose faeces were intact (oocyst shells) constituted 38%) and conformed to morphological features of viable and infectious inoculum oocysts. The fluorescence scores of the inoculated oocysts, obtained by use of the MERIFLUOR test, were identical to those obtained for the faeces-recovered oocysts (majority 3+ to 4+). The dynamics of oocyst shedding showed that overall, the birds released a significantly higher number of intact oocysts than oocyst (P < 0.01). Retention of the viability and infectivity of C. parvum oocysts following intestinal passage through a migratory water-fowl species has serious epidemiological implications. Water-fowl can serve as mechanical vectors for the water-borne oocysts and can contaminate surface waters with C. parvum. As the concentration of Cryptosporidium oocysts in source waters is attributable to water-shed management practices, water-shed protection programme officials should consider water-fowl as a potential factor enhancing contamination of the source water with Cryptosporidium.