Background The aim of this retrospective study was to evaluate the accuracy of bone SPECT (single photon emission computed tomography)/CT (computed tomography) in diagnosing loosening of fixation material in patients with recurrent or persistent back pain that underwent lumbar arthrodesis with pedicle screws using surgery and clinical follow-up as gold standard Methods A total of 48 patients (median age 49 years, range 21-81 years; 17 men) who had undergone lumbar spinal arthrodesis were included in this retrospective analysis. SPECT/ CT results were compared to the gold standard of surgical evaluation or clinical follow-up. Positive SPECT/CT results were considered true positives if findings were confirmed by surgery or if clinical and other examinations were completely consistent with the positive SPECT/CT finding. They were considered false positives if surgical evaluation did not find any loose pedicle screws or if symptoms subsided with non-surgical therapy. Negative SPECT/CT scans were considered true negatives if symptoms either improved without surgical intervention or remained stable over a minimum follow-up period of 6 months. Negative SPECT/CT scans were determined to be false negatives if surgery was still required and loosening of material was found.Results The median length of time from primary surgery to bone SPECT/CT referral was 29.5 months (range 12-192 months). Median follow-up was 18 months (range 6-57) for subjects who did not undergo surgery. Thirteen of the 48 patients were found to be positive for loosening on bone SPECT/CT. Surgical evaluation (8 patients) and clinical follow-up (5 patients) showed that bone SPECT/CT correctly predicted loosening in 9 of 13 patients, while it falsely diagnosed loosening in 4 patients. Of 35 negative bone SPECT/CT scans, 12 were surgically confirmed. In 18 patients, bone SPECT/CT revealed lesions that could provide an alternative explanation for the symptoms of pain (active facet degeneration in 14 patients, and disc and sacroiliac osteodegeneration in 7 patients and 1 patient, respectively). Overall sensitivity and specificity for the detection of loosening were 100 % and 89.7 %, respectively. The positive and negative predictive values were 69 % and 100 %, respectively. Conclusions This retrospective analysis suggests that bone SPECT/CT bone is a highly sensitive and specific tool for the exclusion of screw loosening in patients who present with recurrent low back pain after having undergone lumbar arthrodesis. In addition, it can identify other potential causes of recurrent low back pain in this patient population.