A semi-empirical model was developed to predict simultaneously time-temperature distribution inside the unripe banana and fungal reduction during the hot water dip, in order to simulate different process conditions for controlling banana anthracnose disease. For this purpose, the kinetic parameters of Colletotrichum musae reduction were estimated based on the thermal death time concept and evaluated in different simulated media (unripe and fully ripe banana peel agar media). At the same time, a heat transfer model based on the Fourier equation was developed considering two unripe banana domains (peel and pulp). The comparison between the predicted and observed number of C. musae spores after the thermal treatment of unripe banana was satisfactory (coefficient of variation less than 10%), suggesting that the survival ratio is affected by the banana composition. The resulting model can be used to simulate different time-temperature scenarios, and to show how the parameter combinations may affect C. musae reduction in the unripe banana during the hot water dip process.