Hemangioma is one of the most common benign tumors in newborns. It is a neoplastic proliferative tumor characterized by a period of growth (proliferation) after birth and possible spontaneous involution (regression). Despite the increased interest in the problems of hemangioma treatment and a large number of developed techniques, no clear criteria for choosing a therapy method and its effectiveness exist. Further search for effective treatment methods is necessary because of the variety of forms, localization and prevalence of hemangiomas. The purpose of this work was to develop and test models for predicting the effectiveness of hemangioma treatment in children. Material and methods. The indicators of 84 children with hemangioma who received three types of treatment were used to develop the models. Propranolol was used for the treatment of the first group of children (17 people), timolol was used in the second group (15 people), timolol with compression was used in the third group (52 children). Results and discussion. Hemangioma treatment outcome prediction models have been developed using fuzzy logic. The value of the developed models lies in the fact that indicators that are determined during the initial examination of the child are used to determine the possible outcome of treatment and to choose its optimal tactics. To predict the effectiveness of propranolol treatment the term of the treatment onset and the activity of hemangioma according to hemangioma severity scale are informative. Additional indicators are the period from the birth of a child to the appearance of hemangioma and deviations from the normal values of the clinical blood test indicators. When using timolol, the main informative indicators for predicting the effectiveness of treatment are the period from the treatment onset, the activity of hemangioma according to hemangioma severity scale and the duration of treatment, which can be selected from the interval from 5 to 11 months. The sFasL1 values and the baby’s birth weight are used as additional indicators. When using timolol with compression for predicting the effectiveness of treatment, the main informative indicators are the duration of treatment, the period of the treatment onset and the indicator according to hemangioma severity scale at the beginning of treatment. The duration of treatment can be chosen in each case, based on the desired result, from the range of 5-19 months. Data on the morphological type of hemangioma, the depth of its spread and the presence of factors provoking its growth are used as additional indicators. The average values of errors of the models of predicting indicators characterizing the effectiveness of treatment using various methods were 0.03-9.1% for the indicator according to HSS; for the indicator according to VAS1(visual analog scale) was 2.2-9.4%; for the indicator according to VAS2 was 0.98-5.0%. Conclusion. Developed models for predicting the effectiveness of hemangioma treatment allowed the doctor choosing the most effective method and determining its duration. The software module that implements a support system for a doctor’s decision making as for selecting the optimal method for treating hemangioma can be used in healthcare institutions of various levels of provision of medical aid