There are various factors affecting the fracture of NiTi rotary files. This study was performed to evaluate the effect of cross sectional area, pecking motion and pecking distance on the cyclic fatigue fracture of different NiTi files. Five different NiTi files-Profile®(Maillefer, Ballaigue, Switzerland), ProTaper™ (Maillefer, Ballaigue, Switzerland), K3® (SybronEndo, Orange, CA), Hero 642® (Micro-mega, Besancon, France), Hero Shaper®(Micro-mega, Besancon, France)-were used. Each file was embedded in temporary resin, sectioned horizontally and observed with scanning electron microscope. The ratio of cross-sectional area to the circumscribed circle was calculated. Special device was fabricated to simulate the cyclic fatigue fracture of NiTi file in the curved canal,. On this device, NiTi files were rotated (300rpm) with different pecking distances (3 mm or 6 mm) and with different motions (static motion or dynamic pecking motion). Time until fracture occurs was measured. The results demonstrated that cross-sectional area didn't have any effect on the time of file fracture. Among the files, Profile® took the longest time to be fractured. Between the pecking motions, dynamic motion took the longer time to be fractured than static motion. There was no significant difference between the pecking distances with dynamic motion, however with static motion, the longer time was taken at 3mm distance. In this study, we could suggest that dynamic pecking motion would lengthen the time for NiTi file to be fractured from cyclic fatigue.