Objective: This study investigated the effect of the condition of lubrication on the friction between brackets and NiTi archwires of different rounded cross-sections. Methods: Brackets (Roth, GAC) were affixed to a device connected to a universal testing machine into which segments of archwire were placed (NiTi, Nitinol, GAC) with cross-sections of 0.012-in, 0.016-in and 0.020-in. Once the wire was in the bracket slot, the following lubricants were applied: human saliva (HS: positive control), distilled water (DI), mucin-based (MUC) or carboxymethylcellulose-based (CMC) artificial saliva. In the negative control group, no lubricant was used. The combination between the wire cross-sections and the lubrication condition generated 15 groups with 15 samples each. Data were submitted to two-way analysis of variance and Tukey’s test. Results: There was no significant interaction between the wire cross-section and the condition of lubrication (p= 0.901). Irrespective of whether lubricants were used or not, there was a significant increase in friction with an increase in the cross-section of the wire (p< 0.001). For any wire, the group tested in the presence of MUC was not different from that in which HS was applied. On the other hand, when the application of lubricants was suppressed, significantly higher friction values were observed. The CMC group and the DI group demonstrated intermediate behavior. Conclusions: Friction increased with the increase of the cross-section of the NiTi archwire, but regardless of the archwire, friction with MUC artificial saliva was similar to that of HS and lower than in dry conditions.