OBJECTIVE This research determined whether there is a measurable effect of local geometry factors on leucite content of dental porcelain in fixed partial dentures (FPD). METHODS Four-unit FPD frameworks (n = 36) were fabricated using a nickel-chromium alloy (Rexillium III). Body porcelain (Crystar, shade A2) was applied in one increment and subjected to two simulated body firings, followed by a simulated glaze firing to achieve a thickness of 1.5 to 2.0 mm. The completed FPD specimens were randomly assigned to three groups of 12 specimens each: (1) simulated post-soldering, (2) multiple firing, and (3) control. The FPDs from each test group were sectioned into individual units: canine retainer, premolar pontic, molar pontic, and molar retainer. The porcelain was removed from each unit, and the leucite content was measured via quantitative x-ray diffraction. RESULTS Porcelain cracking indicated that the soldering simulation had successfully reproduced conditions in the dental laboratory that result in porcelain cracking during soldering. The leucite content was not significantly different between the retainer and pontic units for either the soldering simulation or control FPDs, although the canine retainer units did have a slightly lower leucite content than the pooled values of the other units. Comparison of the pooled data for the three groups indicated statistically significant differences among the leucite contents. CONCLUSIONS Compared to the control, the simulated post-soldering procedure produced a significant increase in leucite, and the multiple firing group exhibited a significant decrease in leucite. Increases in leucite weight fraction during post-soldering operations-and the larger thermally induced stresses that accompany these leucite increases-are responsible for the cracking that occurs.