The aim of this study was to evaluate the root surface defect produced by hand curettes and ultrasonic tips with different power settings. Forty root surfaces were divided into 4 groups according the treatment: Gracey curettes, ultrasonic scaler at 10% power, ultrasonic scaler at 50% power and ultrasonic scaler at 100% power. Each specimen was instrumented with 15 strokes and the and divided in the middle to evaluate: (1) the defect depth produced by the instrumentation and (2) contact area of the instrument tips, which was analyzed by scanning electron microscopy. ANOVA and Tukey's test were used for statistical analysis (a=0.05). The results (mean ± SD) of the contact area showed significantly greater defects (p<0.05) for the hand instrumented groups (2092.9 ± 482) compared to the ultrasonic groups (606.8 ± 283.0; 858.6 ± 422.5; 1212.0 ± 366.7, respectively), independently of the power setting. The values for the defect depth on root surface showed no statistically significant difference (p<0.05) between hand instrumentation (66.1 ± 34.0) and ultrasonic scaling at 10%, 50% or 100% power settings (52.4 ± 22.1; 72.0 ± 29.9; 77.7 ± 37.7, respectively). The findings of this study demonstrate that ultrasonic instrumentation produced a similar defect depth to that of hand instrumentation, with a smaller tip contact area, independently of the power setting used for scaling.