The purpose of this study was to investigate the reliability and validity of the Turkish version of the Bath Ankylosing Spondylitis (AS) Functional Index (BASFI). The Turkish version of the BASFI was obtained after a process of translation and back-translation. Eighty-one consecutive patients meeting the 1984 New York criteria for AS were enrolled. Patients were evaluated and requested to complete the questionnaire at days 1 and 2 and on a third occasion between days 15-90. Reliability, reproducibility, validity and sensitivity to change of the Turkish version of the index were assessed. Each score correlated closely with the index score, with coefficients between 0.727 and 0.844. Reliability analysis showed a Cronbach's alpha score of 0.926. Correlations were found between all items of the BASFI and Schober's test (r=-0.258 to -0.531, p<0.001-0.05), occiput-to-wall distance (r=0.284 and 0.589, p<0.001-0.05), and finger-to-floor distance (r=0.334 to 0.613, p<0.001-0.01). The total index score was correlated with the number of nocturnal awakenings (r=0.515, p<0.001), Schober's test (r=-0.444, p<0.001), finger-to-floor distance (r=0.567, p<0.001), occiput-to-wall distance (r=0.535, p<0.001), chest expansion (r=-0.403, p<0.001), and the Dougados articular index (r=0.371, p<0.01). A good correlation was found between day 0 and 1 BASFI indices (r=0.765-0.917, p<0.001), showing good reproducibility of the index. The Turkish version of the BASFI showed reliability, reproducibility, and validity, confirming its utility in the research of AS in Turkey. However, sensitivity to changes due to drug therapy and/or rehabilitation remains to be determined.