Objective: To compare the differences and indications of three evaluation methods for fitness evaluation of removable partial denture (RPD). Methods: A RPD was fabricated and seated on the stone cast of a partially edentulous mandible, and the spaces between RPD and stone cast were recorded with polyvinyl siloxane (PVS) impression material forming PVS replicas. Using cross sectional measurement, the average thicknesses of PVS replicas were measured under stereomicroscope with different numbers of selected measuring points in the denture base, major connector, occlusal rest of the RPD, and the average thicknesses of the PVS replicas measured with different numbers of measuring points were compared using one-way analysis of variance (ANOVA) and independent sample t test. Three kinds of method, including cross sectional measurement, three-dimensional analysis on the stone cast, and three-dimensional analysis on the polyether cast, were applied to measure the average thicknesses of the PVS replicas, and the average thicknesses of the PVS replicas measured by these three evaluation methods were compared with ANOVA. Results: For cross sectional measurement, statistically significant differences were found in the average thicknesses of the PVS replicas in the denture base and the major connector among the different numbers of measuring points (P<0.05), but no differences were found in the average thicknesses of the PVS replicas in the occlusal rest (P>0.05). There were significant differences among the average thicknesses of the PVS replicas measured by these three evaluation methods in each component of the RPD (P<0.01). The average thickness measured by three-dimensional analysis on the stone cast and three-dimensional analysis on polyether cast were smaller than that measured by cross sectional measurement (P<0.05). And there were no differences between the average thicknesses of PVS replicas measured by three-dimensional analysis on stone cast and three-dimensional analysis on polyether cast (P>0.05). Conclusion: For cross sectional measurement, the average thickness of the PVS replicas was influenced by the number of measuring points, and the measurement accuracy of cross sectional measurement was not reliable enough. Three-dimensional analysis on stone cast which is suitable for evaluation in vitro and three-dimensional analysis on polyether cast which is suitable for evaluation in vivo can evaluate the fitness of RPD more comprehensively and effectively than that of cross sectional measurement.