OBJECTIVE the primary aim of this research was to measure fracture toughness for several groups of dental ceramics, and determine how this property is affected by chemistry and microstructure. METHODS Fracture toughness (KIc) values were obtained using Single Edge Precracked Beam (SEPB) and Single Edge V-Notch Beam (SEVNB) methods. Dynamic Young's modulus, which often scales with strength and has been used in explaining the microstructure/toughness relationship on a theoretical basis, was also obtained for the three groups of materials comprising this study. The first group, consisting of micaceous glass ceramics, included model materials that varied systematically in microstructure but not in chemistry. The second group, the feldspathic porcelains, varied significantly in microstructure, but little in chemistry. The ceramics comprising the third group were significantly different in both chemistry and microstructure. RESULTS Upper toughness limits for the micaceous glass-ceramics and feldspathic porcelains were significantly raised compared to the base glasses, but remained under 2 MPa m(1/2). The highest toughnesses were associated with high percent crystallinity, large grains and high aspect ratios. The third group KIc values were 2.8 MPa m(1/2) for a lithium disilicate glass-ceramic, 3.1 MPa m(1/2) for a glass-infused alumina, and 4.9 MPa m(1/2) for zirconia. CONCLUSIONS the correlations between microstructural characteristics and measured properties supports theoretical predictions in the literature. From a practical standpoint, microstructural effects were found to be important, but only within a limited range; the chemistry apparently defined a band of achievable property values. This suggests very large increases in fracture toughness are unlikely to be attained by changes in microstructure alone. A functional relationship determined for the micaceous glass-ceramics enables quantitative predictions of fracture toughness based on the microstructure.