Aim: To compare the adaptation of thermo-plasticised root canal fillings performed in oval canals instrumented with three file systems, and its potential association with debris remaining in recesses. Material and Methods: Thirty-nine mandibular incisors with oval canals were instrumented with WaveOne, ProTaper Universal files or with the self-adjusting file (SAF) system (n = 13). Root canal fillings were performed with Thermafil, with no sealer. The roots were then cut at 2.0, 3.5, 5.0, 6.5 and 8.0 mm from the apex. Morphometric measurements established the percentage of canal area containing no root canal filling material (ANFM). The presence of debris, as a potential cause for defective adaptation, was evaluated microscopically. ANOVA with repeated measures and the Chi-square test were used; significance was set at P 0.05. Results: The apical, round part of the canal showed effective filling with no difference between the file systems. In the oval part of the canal, ≥ 3.5 mm from the apex, the WaveOne and ProTaper Universal groups presented means of 15.2 (± 12.4)% and 12.1 (± 7.5)% of ANFM, respectively, and did not differ from each other. The SAF group had a mean of 6.4 (± 7.9)% of ANFM, which differed significantly from the other two groups (P = 0.001). Deficiencies in filling were associated with the presence of debris in un-instrumented recesses. Conclusions: The adaptation of thermo-plasticised gutta-percha in oval canals was affected by debris remaining or packed into un-instrumented canal recesses. Using the SAF system was associated with less debris and with better adaptation, as compared with the rotary or reciprocating files. Schlagwörter: debris, ProTaper Universal, oval canal, self-adjusting file, thermo-plasticised root fillings, WaveOne