A model predicting the overwintering survival and spring emergence profiles of boll weevil (Anthonomus grandis Boheman) was revised and tested against three years of independent field data from Stonewall County, Texas. The Spring emergence profile was well described by the degree-day-based model, in which the 50 and 80% points of emergence, and the overwintering survivorship varied according to two indices of winter severity. 1988 was an exception, in which the model was 7-12 days early in its predictions. Because the model was not process-based, the hypothesis could not be tested that overwintering attrition causes the widely variable spring emergence profiles. Therefore, an object-oriented, individual-level model has been built based on the first model and preliminary results indicate that the same processes affecting overwintering mortality are responsible for emergence timing differences.