Certain variables were tested for their influence on the color of ceramic metal restorations. The variables included firing temperature, condensation technique, modeling liquid, and brand of porcelain. In addition, the color of different nominal shades was compared. The color was measured with a Minolta CR-121 small-area colorimeter. The small (7 mm2) measuring area enabled readings to be made on actual crown specimens rather than disks. The results were expressed in the CIELAB color system enabling comparisons to be made related to visual perception. The following conclusions can be drawn from the study. 1. The small-area colorimeter is able to detect statistically significant and perceivable color differences between shades of porcelain. The variation in color parameters associated with making replications of a restoration is significantly greater than the error associated with making repeated color measurements of the same restoration. 2. Restorations made with the different brands of porcelain studied have noticeably different colors despite having the same nominal shade. 3. Color changes caused by the choice of modeling liquid were not statistically significant in this study. If the mean color differences are truly representative, using Rainbow or Carv-eze modeling liquids with Vita VMK porcelain may produce color changes that are just barely detectable. 4. The manipulative variables of firing temperature and condensation have little influence on the color of the restorations.