2606 Background: Standard chemotherapy for melanoma is DTIC (RR ~10%). Many physicians do not refer to phase I due to perceived limited clinical benefit (CB=CR+PR+SD) and increased toxicity. To understand the actual experience of melanoma patients (pts) in phase I trials, we analyzed the outcomes of melanoma pts treated on CTEP phase 1 trials (1995-2011) and compared them to DTIC. Methods: We queried the CTMS of CTEP for phase I trials in which advanced melanoma pts were treated. Trials were separated into targeted (T), chemo (C) and immunotherapy (I). Pt characteristics, response and toxicity data were collected. Chemotherapy included chemo with targeted or immunotherapy. Toxicity was drug related if attributed possibly, probably or definitely to drug. Fisher’s Exact Test (2-sided p) was used to compare groups. DTIC data was pooled from 6 modern phase III clinical trials (1999-2011). Results: 937 pts (M595:F342) participated in 148 trials (T: 68, C: 53, I: 27). Characteristics included (median) Age: 51.5 yrs; ECOG status: 0; Prior therapies: 2 (majority receiving prior DTIC); LDH: 206 and albumin: 4.1. Response and toxicity data are shown in the Table. Targeted therapy was associated with lower RR (p=.01), immuno with lower CB rate (p<.001) and chemo with higher incidence of G4 toxicity (p<.001) relative to the other groups. Comparing phase 1 to DTIC, RR and CB were not clinically different (phase I: 6.3% and 26.8% vs. DTIC: 8.8% and 27.9%) but G3-4 toxicity was significantly higher (54% vs. 28%) in phase I (p<.0001). Conclusions: Melanoma pts in prior CTEP phase I trials, a majority DTIC pre-treated, had similar therapeutic benefit but more toxicity compared to DTIC naïve pts in modern clinical trials. [Table: see text]