Background: Immature DC can be loaded in vivo with tumor-specific antigens after tissue destruction and induce specific immune response even in pretreated patients with unfavorable prognosis. We assess clinical efficacy of this approach in melanoma patients after at least 1 line of drug therapy in metastatic setting. Methods: Totally, 14 pts were enrolled: 11 men, 3 women. Stage M1a in 4 (28%) pts, M1b in 1 (7%) and M1c was in 9 (64%) pts. Patient received Cy 300 mg IM at D0, photoditasine (PH) at D3, PhDT therapy on injectable lesion 2 hrs after PH (662 nm, 300 J). The immature DC derived from peripheral blood stem-cells were injected in irradiated lesion 6 hrs after photodynamic therapy at D3 and daily until D7 of the 21-day cycle. Vaccine dose was 60-100*10^6 cells. Toxicity was measured by CTC AE v4, efficacy by RECIST 1.0. Primary endpoint was clinical benefit rate (CBR), secondary – overall survival (OS), response rate (RR), time to progression (TTP) and toxicity. Results: 14 pts were evaluable for toxicity, 13 for response. 37 cycles of therapy were performed. 2 patients achieved PR lasting 393 and 218+ days. 5 patients were stable for 33+, 108, 113, 168 and 239 days. RR was 15 (95% CI 2-39)%, CBR – 54 (95% CI 27-81)%. Median OS was 334 days, TTP – 113 days. RR was well correlated with OS (p=0,014). No SAE or AE of 4^th degree occurred. The only grade 3 toxicity was fever. All grade 1-2 AEs were immune-related: fever (the most frequent AE), pain in the lesions, flu-like reactions, myalgia as well as high serum ALT, AST and bilirubin in patients with liver metastases. Conclusions: Thus, using the described approach for immunotherapy of melanoma has clinical efficacy and worth further developing.