Purpose Literature has described a risk for subsequent vertical subluxation (VS) and subaxial subluxation (SAS) following atlantoaxial subluxation in rheumatoid patients; however, the interaction of each subluxation and the radiographic findings for atlantoaxial fixation has not been described. The purpose of this study was to evaluate the effects of two different posterior atlantoaxial screw fixation on the development of subluxation in patients with rheumatoid atlantoaxial subluxation. Methods Between 1996 and 2006, rheumatoid patients treated with transarticular fixation and posterior wiring (TA) or C1 lateral mass-C2 pedicle screw fixations (SR) in the Nagoya Spine Group hospitals, a multicenter cooperative study group, were included in this study. VS, SAS, craniocervical sagittal alignment, and range of motion (ROM) at the atlantoaxial adjacent segments were investigated to determine whether posterior atlantoaxial screw fixation is a prophylactic or a risk factor for the development of VS and SAS. Results The mean follow-up was 7.2 years (4-12). No statistically significant difference was observed among the patients treated with either of the procedure during the follow-up period. Of 34 patients who underwent posterior atlantoaxial screw fixation, SAS was observed in 26.5 % during the follow-up period; however, VS was not observed. Postoperative C2-7 angle, and Oc-C1 and C2-3 ROM were significantly different between patients with and without postoperative SAS. The incidence of SAS was 38.9 % for TA and 12.5 % for SR; statistically significant differences were observed in the postoperative C1-2 and C2-7 angles, and C2-3 ROM. Conclusions Atlantoaxial posterior screw fixation may be an appropriate prophylactic intervention for VS and SAS if the atlantoaxial joint develops bony fusion following physiological alignment. Compared to TA, SR provided optimal atlantoaxial angle and prevented lower adjacent segment degeneration, thereby reducing SAS.