Purpose. To determine major predictors of thermotherapy (TT) success in retinoblastoma (Rb) treatment using multivariate analysis. Material and methods. A retrospective study of 232 patients (295 eyes, 1623 tumors) that were treated by thermotherapy was performed. The median patients' age was 15±14,1 mo. (from 1 to 78 mo.). Tumor thickness varied from 0,3 to 4,5 mm (Me – 0,7±0,4 mm), tumor basal diameter – from 0,5 to 13,4 mm (Mе – 1,5±1,05 mm). Thermotherapy was conducted both by transpupillary and transscleral routes with the following parameters: wavelength – 810 nm, spot size – 1000–1200 microns, power – from 180 to 850 mW (Ме – 400±156 mW), power density – from 15 to 71 W/сm² (Ме – 33±12,5 W/сm²) in application and scanning modes. Results. Multivariate analysis using Kaplan-Meier estimates and Cox regression analysis showed that tumor thickness ≤0,8 mm (p<0,0001), tumor basal diameter ≤1,5 mm (p<0,0001), preequatorial and postequatorial localization (p<0,0001), dark and medium fundus pigmentation (p=0,0001) and number of treatment sessions ≤2 (p<0,0001) are associated with higher thermotherapy efficacy. ROC-analysis demonstrated that the highest rate of tumor control is associated with three and more predictors of thermotherapy success. Conclusion. Laser thermotherapy proved to be a highly effective method of Rb treatment with 92% rate of tumor control, meanwhile thermotherapy is most effective for tumors that have three or more predictors of thermotherapy success. This should be considered in application of thermotherapy and local modality choice. Key words: laser thermotherapy, retinoblastoma, predictors of efficacy, multivariate analysis, local treatment