Salt crystallization is one of the most important weathering agents and may limit the durability of building stones. Salt crystallization induces stresses inside the pores of stones. Consequently, stone durability is closely related to its physical properties and strength. The purpose of this study was to propose a statistical model for estimating stone durability against salt crystallization considering both the physical properties and strength of the stones -utilizing multiple regression. For this purpose, 14 samples of building stones were
selected and their mineralogical characteristic, physical properties, and strength (density, porosity, water absorption, uniaxial compressive strength, and Brazilian tensile strength) were determined. Then, the salt crystallization test at a sodium sulfate solution of up to 20 cycles was carried out, and the Dry Weight Loss (DWL) of samples was measured. The durability of each sample was assessed by the percentage of weight loss after the salt crystallization test. The relationships between stone durability and the physical properties and strength of the stones- using simple and multiple regression analyses- were investigated. Moreover, statistical models for estimating stone durability were proposed. These models show that stone durability can be estimated accurately by considering both the physical properties and strength characteristics.