A small-volume sentinel chamber was developed to assess the effects of environmental stresses on survival of sucrose-Percoll-purified Cryptosporidium parvum oocysts in soil and animal wastes. Chambers were tested for their ability to equilibrate with external chemical and moisture conditions. Sentinel oocysts were then exposed to stresses of the external environment that affected their viability (potential infectivity), as indicated by results of a dye permeability assay. Preliminary laboratory experiments indicated that temperatures between 35 and 50°C and decreases in soil water potential (0.003 to 3.20 MPa) increased oocyst inactivation rates. The effects of two common animal waste management practices on oocyst survival were investigated on three dairy farms in Delaware County, N.Y., within the New York City watershed: (i) piling wastes from dairy youngstock (including neonatal calves) and (ii) spreading wastes as a soil amendment on an agricultural field. Sentinel containers filled with air-dried and sieved (2-mm mesh) youngstock waste or field soil were wetted and inoculated with 2 million oocysts in an aqueous suspension and then placed in waste piles on two different farms and in soil within a cropped field on one farm. Controls consisted of purified oocysts in either phosphate-buffered saline or distilled water contained in sealed microcentrifuge tubes. Two microdata loggers recorded the ambient temperature at each field site. Sentinel experiments were conducted during the fall and winter (1996 to 1997) and winter (1998). Sentinel containers and controls were removed at 2- to 4-week intervals, and oocysts were extracted and tested by the dye permeability assay. The proportions of potentially infective oocysts exposed to the soil and waste pile material decreased more rapidly than their counterpart controls exposed to buffer or water, indicating that factors other than temperature affected oocyst inactivation in the waste piles and soil. The effect of soil freeze-thaw cycles was evident in the large proportion of empty sentinel oocysts. The potentially infective sentinel oocysts were reduced to <1% while the proportions in controls did not decrease below 50% potentially infective during the first field experiment. Microscopic observations of empty oocyst fragments indicated that abrasive effects of soil particles were a factor in oocyst inactivation. A similar pattern was observed in a second field experiment at the same site.