The aim of the study was to conduct a long-term follow-up on the stability of the hard tissues after bilateral sagittal split osteotomy (BSSO) with rigid internal fixation (RIF)to set back the mandible and to compare it with that of mandibular advancement performed by the same team of surgeons and with the same examination protocol. Seventeen consecutive patients (6 females and 11 males) could be re-examined 12.7 years (T5) after surgery. The previous examinations were before surgery (T1), 5 days (T2), and 6.6 (T3) and 14.4 (T4) months after surgery. Lateral cephalograms were traced by hand, digitized, and evaluated with the Dentofacial Planner software program. The x-axis for the system of co-ordinates ran through sella (point zero) and the line nasion-sella-line minus 7 degrees. The program determined the x- and y-values of each variable and the usual angles and distances. The effects of treatment were determined with Wilcoxon matched pairs, signed ranks test, with Bonferroni adjustment, and the relationship between variables with Spearman rank correlation coefficient. Relapse at point B was 0.94 mm or 15 per cent and at pogonion 1.46 mm or 21 per cent of the initial setback at T5. Relapse was mainly short-term (T4-T2), 13 per cent for point B and 17 per cent for pogonion. Gender correlated significantly with relapse (T5-T2) at point B (P = 0.002) and pogonion (P = 0.021), i.e. females in contrast to males showed further distalization of the mandible instead of relapse. No correlations were seen for age or the amount of surgical setback. The long-term results in mandibular setback patients were more stable when compared with the mandibular advancement patients examined previously. The initial soft tissue profile, the initial growth direction, and the remodelling processes of the hard tissues must be considered as reasons for long-term relapse. Growth direction positively influenced the long-term results in females: further distalization of the mandible occurred.