Models have been developed to estimate leaf wetness duration (LWD) using measured or estimated weather data on the basis of approaches such as energy balance equations, neural networks, and classification and regression trees (CART). Models that embody physical principles ensure spatial portability but usually require accurate and extensive input data to estimate LWD accurately. Empirical models may be more tolerant to errors of input data and require more limited weather inputs, but they rarely possess wide portability because they do not incorporate physical principles. In this study, a hybrid model was developed to incorporate both energy balance principles and empirical approaches by using fuzzy logic. The results suggested that a LWD model based on a fuzzy logic system offers advantages in comparison to the previous models since the model possessed wider portability than strictly empirical models. Empirical methodologies included in the model algorithm allowed a relatively small number of input variables and tolerated imprecise weather data input. The fuzzy LWD model also possessed adaptability to specific circumstances using a correction factor, which can be determined through a simple training process. For example, when LWD was predicted with site-specific weather forecasts in which substantial systematic errors are contained, the fuzzy LWD model was able to forecast LWD accurately using a correction factor. The correction factor also expanded spatial portability of the fuzzy LWD model to environments in which climate conditions differed considerably, e.g., from temperate to tropical zones. The fuzzy LWD model, therefore, deserves further attention as a substitute for current physical and empirical LWD models.