We used a gene amplification strategy to analyze T-cell receptor (TCR) gene rearrangements in 185 specimens, including mycosis fungoides/Sezary syndrome (MF/SS), other cutaneous neoplasms, inflammatory dermatoses, reactive lymphoid tissues, and normal skin. Genomic DNA was extracted from lesional tissues and rearrangements of the TCR-gamma chain gene were amplified using the polymerase chain reaction (PCR) with primers specific for rearrangements involving V gamma 1-8 or V gamma 9 gene segments. The resulting PCR products were then separated according to their nucleotide sequence as well as size by denaturing gradient gel electrophoresis (DGGE). Dominant clonal TCR-gamma gene rearrangements were detected in 61 of 68 MF/SS cases by PCR/DGGE. This sensitivity of 90% compared to a sensitivity of only 59% when dominant clonality was sought in 17 of these same cases by Southern blot analysis of TCR-beta gene rearrangements. This difference in sensitivity was greatest in early, minimally infiltrated skin lesions. PCR/DGGE was also more sensitive than Southern blot analysis for detecting peripheral blood involvement in two cases of early MF. Among 12 additional specimens of suspected MF/SS, nine (75%) showed clonal TCR-gamma gene rearrangements by PCR/DGGE including six of eight cases with a previously confirmed diagnosis of MF/SS and three of four cases without prior known MF/SS. Among 105 non-MF/SS specimens, dominant TCR-gamma gene rearrangements were detected in only six cases (6%). Four were diagnosed as chronic dermatitis and two were diagnosed as cutaneous lymphoid hyperplasia. We conclude that the large majority of MF/SS cases, including patch phase disease, possess dominant clonal TCR-gamma gene rearrangements. PCR/DGGE is more sensitive than Southern blot analysis for detecting dominant clonality and staging disease in patients with a confirmed diagnosis of MF/SS. However, because PCR/DGGE is sensitive enough to detect dominant TCR-gamma gene rearrangements in a subset of patients with chronic dermatitis, it cannot be used as the sole criterion for establishing a diagnosis of T-cell lymphoma. As with other molecular biologic clonality assays, clinicopathologic correlation is essential. Nevertheless, the detection of dominant clonality in some cases of histologically nonspecific dermatitis allows the identification of a previously unrecognized subset of patients, i.e., those with "clonal dermatitis." It will be important to determine the long-term risk of MF/SS among these patients because our study indicated that MF/SS can sometimes present with lesions indistinguishable from clonal dermatitis.