Lake trout Salvelinus namaycush, aged 3 and 6 years and with average weights of 700 and 2,000 g, were grown in laboratory tanks for up to 407 d under a thermal regime similar to that experienced by lake trout in nearshore Lake Michigan. Lake trout were fed alewifeAlosa pseudoharengus and rainbow smelt Osmerus mordax, prey typical of lake trout in Lake Michigan. Of the 120 lake trout used in the experiment, 40 were fed a low ration (0.25% of their body weight per day), 40 were fed a medium ration (0.5% of their body weight per day), and 40 were fed a high ration (ad libitum). We measured consumption and growth, and we compared observed consumption with that predicted by the Wisconsin bioenergetics model. For lake trout fed the medium ration, model predictions for monthly consumption were unbiased. Moreover, predicted cumulative consumption by medium-ration lake trout for the entire experiment (320 d for smaller lake trout and 407 d for larger lake trout) agreed quite well with observed cumulative consumption; predictions were as close as within 0.1 to 5.2% of observed cumulative consumption. Even so, the model consistently overestimated consumption by low-ration fish and underestimated consumption by high-ration fish. The bias was significant in both cases, but was more severe for the low-ration trout. Because the low-ration and high-ration regimes were probably unrealistic for lake trout residing in Lake Michigan and because the model fit our laboratory data rather well for medium-ration trout, we conclude that applying the Wisconsin bioenergetics model to the Lake Michigan lake trout population in order to estimate the amount of prey fish consumed by lake trout each year is appropriate.