The molecular basis of parathyroid adenomatosis includes defects in the cyclin D1/PRAD1 and MEN1 genes but is, in large part, unknown. To identify new locations of parathyroid oncogenes or tumor suppressor genes, and to further establish the importance of DNA losses described by molecular allelotyping, we performed comparative genomic hybridization (CGH) on a panel of 53 typical sporadic (nonfamilial) parathyroid adenomas. CGH is a new molecular cytogenetic technique in which the entire tumor genome is screened for chromosomal gains and/or losses. Two abnormalities, not previously described, were found recurrently: gain of chromosome 16p (6 of 53 tumors, or 11%) and gain of chromosome 19p (5 of 53, or 9%). Losses were found frequently on 11p (14 of 53, or 26%), as well as 11q (18 of 53, or 34%). Recurrent losses were also seen on chromosomes 1p, 1q, 6q, 9p, 9q, 13q, and 15q, with frequencies ranging from 8-19%. Twenty-four of the 53 adenomas were also extensively analyzed with polymorphic microsatellite markers for allelic losses, either in this study (11 cases) or previously (13 cases). Molecular allelotyping results were highly concordant with CGH results in these tumors (concordance level of 97.5% for all informative markers/chromosome arms examined). In conclusion, CGH has identified the first two known chromosomal gain defects in parathyroid adenomas, suggesting the existence of direct-acting parathyroid oncogenes on chromosomes 16 and 19. CGH has confirmed the locations of putative parathyroid tumor suppressor genes, also defined by molecular allelotyping, on chromosomes 1p, 6q, 9p, 11q, 13q, and 15q. Finally, CGH has provided new evidence favoring the possibility that distinct parathyroid tumor suppressors exist on 1p and 1q, and has raised the possibility of a parathyroid tumor suppressor gene on 11p, distinct from the MEN1 gene on 11q. CGH can identify recurrent genetic abnormalities in hyperparathyroidism, especially chromosomal gains, that other methods to not detect.