1. Estimating the probability of establishment is an important step toward managing ecological risk posed by natural or intentional introductions. Species introductions in certain geographic areas may pose less of a threat when environmental conditions are unlikely to promote long-term establishment. Species commonly referred to as 'tilapia' have become widespread in certain areas of the southeastern United States, yet concerns remain regarding their potential spread to other areas. 2. We created a model to predict the survival of tilapia in Georgia (U.S.A.) based on individual 'thermal experience'. Laboratory experiments were conducted to measure the duration of survival of two strains of tilapia, Oreochromis nilotica and O. nilotica · O. aureus hybrids, exposed to a variety of temperature regimes with three minimum temperatures and two rates of temperature decline. 3. We used Mayfield hierarchical logistic regression (MHLR) to describe the daily probability of survival as a function of the rate of temperature decline, average sustained temperature, tilapia strain and mass-length ratio. 4. Tilapia generally survived constant minimum temperatures over 12°C but survival rate was mediated by the rate of decline, mass-length ratio and strain. For every 1°C increase in minimum temperature, tilapia were 2.76 times more likely to survive. Fish exposed to rapid temperature decline were less likely to survive than those exposed to slowly decreasing temperature. More robust fish (higher mass-length ratio) were more likely to survive, and strain had a minimal effect despite being supported by the best MHLR model. 5. Our model can be used to estimate survival probability for tilapia under known surface water temperature regimes. Using MLHR in conjunction with experimental tolerance data may be useful for estimating the likelihood of establishment of potentially invasive species.