OBJECTIVES: This research was conducted to demonstrate the effectiveness of reflective photoelasticity as an in vivo technique for monitoring the strain/stress distribution in prosthetic devices during function. METHODS: Seven subjects were selected for this study and divided into three groups according to the design of the prostheses. The buccal surfaces of the experimental prostheses were coated with a bi-refringent resin 2mm thick. Using a reflection polariscope, fringe patterns were documented on video tape during standardized in vivo loadings. Strain and stress levels (gamma(max) and tau(max)) were calculated from isochromatic fringe order determinations. The intra-observer variability in counting fringe orders and the accuracy level of reflective photoelasticity as compared to strain gauge results were also studied. The strain results of in vivo loading of the prostheses were analyzed with ANOVA and Scheffes tests. The intra-observer variability was analyzed with the Friedman nonparametric test. Comparisons of photoelastic data and strain gauges were tested with Pearson's correlation. RESULTS: Coated areas accessible to normally incident light, produced comprehensive maps of strain distribution. Significant differences were found in the in vivo gamma(max) values for the prostheses designs (p<0.05). The test for intra-observer variability showed that no significant differences occurred in counting fringe orders. Strain values obtained with reflective photoelasticity showed a high correlation r=0.98 and 0.99) with values obtained at the same areas using strain gauges. SIGNIFICANCE: Reflective photoelasticity is a valid, reliable and accurate technique to be used for in vivo studies on the biomechanical behavior of prosthetic devices.