Aim. To evaluate changes in the concentration of biomarkers for osteoproliferation and bone resorption in ankylosing spondylitis (AS) patients treated with nonsteroidal anti-inflammatory drugs (NSAIDs) in different regimens. Subjects and methods. Forty patients with AS (according to the modified New York criteria), who had BASDAI ≥ 4.0 at baseline and at 52 weeks of on-demand NSAID treatment were examined and randomized into 2 groups: 1) 30 patients who used continuously oral tenoxicam 20 mg daily (a study group); 2) 10 patients who continued previous therapy (a comparison group). BASDAI and ASDAS were calculated; the serum levels of C-reactive protein, C-terminal type I procollagen propeptide (PICP), and C-terminal telopeptide of type I collagen (CTX-I) were measured at baseline and at 52 and 56 weeks of treatment. A control group consisted of 19 healthy volunteers. Results. The continuous use of NSAIDs (tenoxicam) decreased higher baseline BASDAI and ASDAS scores. There were no changes in the indicators of AS activity in the patients who took on-demand NSAIDs. Baseline CTX-I levels did not differ between the patients with AS and the healthy individuals; those declined during continuous intake of tenoxicam and remained unchanged during on-demand administration. In the patients with AS, baseline PICP levels exceeded those in the healthy individuals. In the tenoxicam-treated patients, the concentrations of PICP at baseline and at 52 and 56 weeks were 17.1±9.0, 16.8±9.9, and 13.29±6.7 ng/ml, respectively (p=0.0001 for differences between the baseline and week 56 levels); in the comparison group, PICP levels did not change statistically significantly (p≥0.05 for all intergroup comparisons). Conclusion. Changing the inefficient long-term on-demand use of NSAIDs to their continuous intake is associated with a rapid decrease in clinical AS activity (within 4 weeks) with a reduction in the higher baseline concentration of the marker for osteoproliferation and in the normal level of the marker for bone resorption.