The whooping crane (Grus americana) breeds in a vast wetland complex in Wood Buffalo National Park (WBNP), Canada. Joint conservation efforts between Canada and the United States, the location of the crane’s winter home, have increased the population from 16 individuals in the 1940’s to 220 birds in 2005. The National Recovery Plan for the Whooping Crane and the Draft Revised International Recovery Plan for the Whooping Crane have outlined the need to monitor the wetland habitat characteristics and whooping crane prey, and relate this information to chick survival and opportunities for population expansion. To develop a prey monitoring program in the breeding grounds, we sampled a total of 45 ponds in 2004 and 2005. We first examined sampling efficiency of timed dip-net sweeps, minnow traps and activity traps. We then assessed spatial variation of fish and invertebrate assemblages by sampling ponds from six nesting areas, and assessed temporal variation by repeatedly sampling ponds (PC ponds) that were similar to nesting area ponds, but more easily accessible. Crane-use of the seven areas was classified a priori as high-, low- or no-use. Taxonomic composition of samples and taxa accumulation curves indicate that a combination of minnow traps and sweeps will monitor prey assemblages most efficiently. Multi-response permutation procedure (MRPP) analyses indicate that prey composition in high-use ponds differed from both low- and no-use ponds, which did not differ. Indicator species analysis showed that dace and stickleback are indicators of high-use nesting areas, though not all high-use area ponds contain fish. MRPP analyses of temporal PC ponds indicated that prey assemblages differed in each time period, with dominance shifting from Dytiscidae and Trichoptera in June, to Odonata in July, to Ephemeroptera in August. Assessment of both temporal and spatial changes in prey must be incorporated in the whooping crane prey monitoring program.