Boll weevil, Anthonomus grandis grandis Boheman, overwintering survivorship was quantified monthly throughout the overwintering period (October to May) in Texas High Plains and Rolling Plains for 12 yr. A negative exponential model was developed to dynamically predict survivorship throughout the overwintering months. Survivorship was modeled as a function of the number of days that weevils were in the habitat, negative degree-days (<0.0°C), positive degree-days (>6.1°C), rainfall, and mortality during the first month of overwintering. First month mortality was modeled as a function of overwintering survival potential of weevils determined by dissection examination of their body lipid content and gonad atrophy. A nonlinear iterative multiple regression analysis showed that the model explained 94% of the variability in parameterization-verification data; a goodness-of-fit test showed that 97% of the estimated survival values did not significantly depart from their corresponding observed values. With independent validation data, 94% of the variability was explained by the survival model; a goodness-of-fit test for validation data showed that 96% of the predicted survival values did not significantly depart from their corresponding observed values. This model offers a greater understanding of boll weevil overwintering biology as it demonstrates a link between biological and climatic parameters. The model can be used to forecast weevil survivorship throughout the overwintering period in the Texas Plains.