The leucite particles in dental porcelains are often partially encircled by microcracks that are the result of the thermal expansion mismatch between leucite and the surrounding glass matrix. Although the magnitude of the stress at the particle-matrix interface is independent of the particle size (Selsing, 1961), Davidge and Green (1968) showed experimentally that there is a critical particle size below which microcracking is absent. The critical particle size is explained by a Griffith-type energy balance criterion: Below the critical size, the stress magnitude may be sufficient to cause cracking, but there is insufficient strain energy for the creation of the new surfaces of the microcrack. The purpose of the present study was to determine whether the mean leucite particle size of a dental porcelain influences the degree of microcracking in the porcelain. Microcrack density, leucite particle surface area per unit volume, and leucite mean volume-surface diameter, D3,2, were determined by quantitative stereology on 10 specimens each of 6 dental porcelains and Component No. 1 of the Weinstein et al. patent (US Patent 3,052,982, 1962). The fraction of leucite particles with microcracks around them, f(mc), was estimated for each porcelain from the microcrack density and the leucite surface area. Using the equations of Selsing (1961) and Davidge and Green (1968), we calculated the critical particle diameter, Dc, for leucite to be 4 microm. The porcelains were partitioned according to whether their mean leucite particle diameters, D3,2, fell above or below Dc, and their values of f(mc) were analyzed by a permutation test with random re-sampling. The porcelains with mean leucite particle diameters below Dc had a significantly lower fraction of cracked particles compared with the porcelains with mean leucite particle diameters above Dc (p < 0.05). This study provides evidence that microcracking in dental porcelain can be minimized by a reduction of the mean leucite particle diameter to less than 4 microm.