PURPOSE: The current American Joint Commission on Cancer (AJCC) staging system distinguishes between soft tissue and visceral metastases in advanced (stage IV) melanoma. We sought to verify these staging criteria and to identify prognostic variables that could be used to evaluate the impact of systemic therapy on long-term survival during the prior decade. PATIENTS AND METHODS: We conducted a retrospective study of patients with advanced cutaneous melanoma enrolled in clinical trials between 1979 and 1989 at The University of Texas M.D. Anderson Cancer Center. Pretreatment age, sex, number of organs with metastases, serum levels of lactate dehydrogenase (LDH) and albumin, and period of enrollment were analyzed using a Cox proportional hazards model of survival. RESULTS: In univariate and multivariate analyses that involved 318 stage IV patients, normal serum levels of LDH and albumin, soft tissue and/or single visceral organ metastases (especially lung), female sex, and enrollment late in the decade were independent positive predictors for survival. In multivariate analyses, the current AJCC criteria did not significantly predict outcome. Systemic treatment response did not bias these results, and only 4% of patients had a complete response. Patients who lived more than 2 years (11%) had a mix of favorable prognostic characteristics and a high frequency of systemic or surgically induced complete response. CONCLUSION: This study supports the use of stratification parameters that reflect the favorable prognostic impact of soft tissue or single visceral organ metastases and normal serum levels of LDH and albumin at time of enrollment in advanced melanoma trials. Improved survival over the prior decade probably reflects advances in diagnostic and palliative interventions.