The histopathologic distinction between pigmented actinic keratosis (PAK) and atypical junctional melanocytic proliferations (AJMP) is a common problem, and it is one with meaningful clinical significance. Previous publications have suggested that Melanocyte Antigen Related to T-cells-1 (MART-1)--a melanocytic marker related to host immune response--was not useful in making this interpretative separation. To revisit that assertion, the authors selected 68 specimens that concerned the diagnosis of PAK vs. AJMP. The degree of morphologic difficulty attached to each case was rated semiquantitatively using a three-tiered scale, and interpretative problems were caused by cytologic similarity between atypical keratinocytes and aberrant melanocytes, obscuring lichenoid inflammation, subepidermal fibrosis, and an absence of clearly defined cell nests at the dermoepidermal junction. Each biopsy sample was immunostained for MART-1 (using antibody clone A103) with azure-B counterstaining; the principal criterion for a diagnosis of AJMP was that of confluent cellular positivity over at least 1 high-power (x400) microscopic field, in conjunction with nested cell growth. The specimens were then re-examined diagnostically. Immunostaining definitely improved interpretative certitude in 65 examples (96% effectiveness); the final diagnosis was that of PAK for 21 lesions and AJMP for 47. Three specimens--all of which represented AJMP--did not benefit by MART-1 staining. It is concluded that MART-1 immunostaining with azure-B counterstaining is a useful adjunct in the interpretation of problematic intra-epidermal pigmented lesions.