The possibility of predicting the resistance of tropical hardwoods to fungal decay was studied. A multiple regression equation for predicting the natural durability of tropical timbers was proposed. The weight loss of 24 commercially popular Malaysian hardwoods was obtained by a modified ASTM D2017 soil block method using 3 white-rot fungi, Coriolus versicolor, Ganoderma lucidum, Pycnoporus coccineus, and a brown-rot fungus, Tyromyces palustris. The percentage of weight loss induced by T. palustris was selected as a criterion variable (Y). Various properties of the woods which could affect the durability of timbers, such as extractives content, lignin content, pH, density, water absorption capacity, and morphological characteristics were analyzed. The factors selected as independent variables based on the significance of coefficients for the weight loss induced by T. palustris were the density (X1), water absorption capacity (X2), methanol extractives content (X3), and pH (X4). The weight loss calculated by the equation was the lowest in Chengal (Neobalanocarpus heimii). The regression equation obtained was Y = 18.546 ? 37.028 (X1) + 0.016 (X2) ? 1.056 (X3) + 8.056 (X4). The durability of the timbers estimated by the equation matched the natural durability obtained from field stake tests with minor differences.