We evaluated the influence of manual preflaring and torque on the failure rate of rotary nickel-titanium ProTaper instruments Shaping 1 (S1), Shaping 2 (S2), Finishing 1 (F1), and Finishing 2 (F2). These factors were evaluated using an in vitro method by calculating the mean number of Endo-Training-Blocks shaped before file breakage under different conditions. Group A (S1 on simulators with no preflaring) shaped 10 blocks before failure, group B (S1 on manually preflared simulators) shaped 59 blocks (p<0.01 versus group A), group C (S2 with low torque) shaped 28 blocks, group D (S2 with high torque) shaped 48 blocks (p<0.01 versus group C), group E (F1 with low torque) shaped eight blocks, group F (F1 with high torque) shaped 23 blocks (p<0.01 versus group E), group G (F2 with low torque) shaped four blocks, and group H (F2 with high torque) shaped 11 blocks (p<0.01 versus group G). Manual preflaring creates a glide path for the instrument tip and is a major determinant in reducing the failure rate of these rotary nickel-titanium files. All instruments worked better at high torque.