Objective: To evaluate the long-term results of local application of transpupillary photodynamic therapy (PDT) with chlorine photosensitizer in isolated organ-preserving treatment of uveal melanoma (UM) and to determine predictors of its effectiveness. Patients and methods. Retrospective analysis of transpupillary PDT efficacy in 38 patients (38 eyes) diagnosed with choroidal melanoma (CM) treated between 2016 and 2024. The median follow-up time was 24 months. The mean age was 62.77 ± 10.89 years. Distribution according to TNM: I (n = 12), IIA (n = 20), IIB (n = 6). According to ultrasound Doppler, the initial tumor thickness before treatment ranged from 1 mm to 5.2 mm, with a median value of 2.95 (2.2; 4.0). Distribution by degree of vascularization: avascular СM (n = 4), hypovascular CM (n = 12), hypervascular CM (n = 22); by degree of pigmentation: weak (n = 22), moderate (n = 11), severe (n = 5). In a multivariate statistical analysis of predictors of transpupillary PDT efficacy, the following features were taken into account: tumor thickness, the degree of pigmentation, and the degree of UM vascularization. Results. Complete regression was achieved in 32 of 38 patients (84 %). After treatment, the tumor thickness ranged from 0 to 4 mm, the median value was 0.5 (0.3; 1.0). Mean best corrected visual acuity before treatment was 0.70 ± 0.34 and 0/38 ± 0.37 after treatment. Multivariate analysis of predictors of treatment efficacy reveal that the most significant criterion was the degree of CM vascularization (AUC = 0.906 (0.802; 1.000), p = 0.002). Complete regression was achieved in 100 % of cases with hypervascular UM, 75 % with hypovascular, and 25% with avascular (p = 0.001). Complete regression was significantly more often observed in small CM (AUC = 0.922 (0.795; 1.000), p = 0.001). The optimal threshold values of the CM thickness for transpupillary PDT in an isolated treatment were obtained: 3.9–4.5 mm. It has been proved that the degree of CM pigmentation is a significant predictor of the absence of complete regression (AUC = 0.805 (0.595; 1.000), p = 0.019), so in the weakly pigmented variant, the formation of an atrophic scar was determined in 21 of 22 patients (95.5 %), in the moderate variant — in 9 of 11 cases (81.8 %), in the case of severe pigmentation in 2 of 5 (40 %) patients (p = 0.012). Conclusions. Transpupillary PDT efficacy in the isolated treatment of small and medium CM is determined by the possibility of achieving complete regression in 2/3 of patients. The identified predictors should be taken into account when conducting this type of treatment. © 2025 Ophthalmology Publishing Group. All rights reserved.