Predation is an important factor in the dynamics of gypsy moth (Lymantria dispar L.) populations, yet predation rates can be difficult to estimate accurately in the field. Biased estimates can result from spatial heterogeneity in risk or from artifacts associated with deploying prey. Here we compare predation rates on freeze-dried gypsy moth pupae affixed with beeswax to pieces of burlap with predation rates on naturally occurring live pupae in the same sites. Daily predation rates, primarily by small mammals, were two to eight times greater for freeze-dried deployed pupae than natural pupae, depending on the year. These results indicate apparent predation rates can be substantially biased by artifacts associated with deployed prey, such as human scent, artificial substrates, or freeze drying. Results from studies using similar methods may provide qualitative comparisons of relative predation risk, but their estimates of absolute predation rates should be interpreted with caution, and attempts should be made to quantify and correct for any resulting bias.