<h2>Abstract</h2> <b>Statement of Problem.</b> Research-based assessments of the quality of removable partial dentures (RPDs), especially within a large patient population, are lacking. <b>Purpose.</b> This study evaluated the prevalence and quality of RPDs by use of the National Health and Nutrition Survey (NHANES III) data set. <b>Material and Methods.</b> Estimates of the health and nutritional status of the American population were obtained from the NHANES III data set (1988-1994). Calibrated dentists performed dental examinations on 17,884 adults. If the subject wore an RPD, its quality was assessed with 5 criteria: integrity, tooth wear, the presence of temporary reline material or adhesive, stability, and retention. In this study, the data on RPD defects were used to assess the prevalence of problems. The original population was divided into 4 subgroups: paired RPD data=subjects wearing both maxillary and mandibular RPDs (n = 600 prostheses, 300 maxillary and 300 mandibular); single RPD data=subjects with only one RPD (n = 1003 prostheses, 511 maxillary and 492 mandibular); maxillary versus mandibular RPD data=all subjects with all RPDs (n = 1603 prostheses, 811 maxillary and 792 mandibular); and pooled RPD data=all subjects with a single RPD plus subjects with 2 RPDs, counting only the RPD with the most defects (n = 1303 prostheses, 674 maxillary and 629 mandibular). The subjects were analyzed with descriptive statistics. The distribution of defects in the RPDs was examined for possible associations with χ<sup>2</sup> tests. The paired data for patients with both maxillary and mandibular RPDs were analyzed with Wilcoxon signed rank tests. The prevalence of RPDs as a function of patient age and the association between RPD defects and patient age were analyzed with χ<sup>2</sup> tests. The results were considered highly significant at <i>P</i><.0001 and significant at <i>P</i><.05. <b>Results.</b> Of the 17,884 adults who underwent a dental examination, 1306 had RPDs. Three patients were excluded because their records were incomplete, leaving 1303 patients available for analysis. Most of the prostheses examined (65%) had at least 1 defect. Lack of stability was the most prevalent single defect. Distinctions in the type and prevalence of defects were observed between mandibular and maxillary RPDs. Mandibular RPDs had significantly more problems related to retention, whereas maxillary RPDs had significantly more problems related to the presence of reline material and to integrity defects. Tooth wear defects were significantly associated with patient age (<i>P</i><.0001). Only one third of the RPDs were considered satisfactory according to NHANES III criteria. <b>Conclusion.</b> A review of the database used in this study indicated that, in spite of a decline in tooth loss, RPDs are still used in all age cohorts, including young adults. A large number of RPDs were found to have defects. (J Prosthet Dent 2002;88:37-43.)