Objective: to analyze the extent of analgesic effect and to determine predictors of inadequate response to local therapy with non-steroidal antiinflammatory drugs (NSAIDs) in a prospective, comparative, randomized trial of the efficacy and safety of Artoxan® gel 1% versus Diclofenac gel 1% in patients with knee OA.Material and methods. The study included 60 patients with a definite diagnosis of stage II–III Kеllgren–Lawrence knee OA who fulfilled ACR criteria and were observed on an outpatient basis in V.A. Nasonova Research Institute of Rheumatology. Patients were 40–80 years old (mean 62.50±8.04 years), body mass index (BMI) 24.9±4.67 kg/m2 , median OA duration 5.7 [3;15] years. According to the randomization scheme, the patients were divided into two groups. In the 1st group (n=30), local therapy with 1% Artoxan gel was applied to the target area of the knee twice daily for 14 days. Patients in the 2nd group (n=30) were prescribed local therapy with the comparator drug, 1% Diclofenac gel with a similar application regimen. Patients in both groups were comparable in terms of the main parameters.Results and discussion. Patients in both groups showed a significant decrease in pain intensity in the target joint during walking according to the visual analogue scale (VAS) after two weeks of treatment (p <0.05). A decrease in pain (to mild or moderate) in the target joint to <40 mm according to VAS after 7 days of therapy reported 43.3% of patients in the 1st group, and 63.3% of patients after 14 days of therapy (p=0.09). In the 2nd group, 43.3 % of patients also reported a reduction in pain in the target joint to <40 mm according to VAS after 7 days of therapy, and after 14 days it was observed in 56.7% of cases (p=0.22). Although the differences between the groups did not reach statistical significance, a reduction in pain to <40 mm according to VAS and a high BMI (r= -0.28; p=0.029).Conclusion. The results of the study demonstrate a significant analgesic effect of local NSAIDs in knee OA. In most patients, pain was <40 mm according to VAS after 2 weeks of local NSAID therapy. At the same time, there was a tendency towards a higher frequency of pain reduction to <40 mm according to VAS in the group receiving local therapy with 1% Artoxan gel. It was concluded that excessive body weight and high BMI may be predictors of inadequate analgesic effect in patients with knee OA.