Three models (linear, third-degree polynomial, and one based on enzyme kinetics) were used to simulate the dependence ofCalosoma sycophantaL. developmental rate on temperature. In conjunction with a model of individual differences in development based on the Weibull distribution, the models were used to predict development ofC. sycophantaat fluctuating temperatures in a greenhouse and in nature. In the greenhouse, the reaction-rate model based on enzyme kinetics was most consistently correct. Simulated development was 1–2 d faster than actual, but this may have been partly because of the large fraction of time greenhouse temperatures were below 20°C. Simulations for one set of field data had to be adjusted for the time of larval hatch, but the same adjustment gave a good simulation of independent data. One or more of the temperature-based models, in conjunction with other studies on food consumption, may prove useful in measuring the effect of this predator on gypsy moths.