BACKGROUND/OBJECTIVE: Exercise-based rehabilitation is a conservative management approach for individuals with low back pain. However, adherence rates for conservative management are often low and the reasons for this are not well described. The objective of this study was to evaluate predictors of adherence and patient-reported reasons for non-adherence after ceasing a supervised exercise-based rehabilitation program in individuals with low back pain. DESIGN: Retrospective observational study. METHODS: Data was retrospectively analyzed from 5 rehabilitation clinics utilizing a standardized exercise-based rehabilitation program. Baseline demographics, diagnosis and symptom specific features, visit number, and discontinuation profiles were quantified for 2,243 patients who underwent the program. RESULTS: Forty-three percent (43%) of participants were adherent to the program, with the majority (31.7%) discontinuing treatment prior to completion due to logistic and accessibility issues. Another 13.2% discontinued prior to the prescribed duration due to clinically significant improvements in pain and/or disability without formal discharge evaluation, whereas 8.3% did not continue due to lack of improvement. Finally, 6.0% were discharged for related and unrelated medical reasons including surgery. Individuals diagnosed with disc pathology were most likely to be adherent to the program. LIMITATIONS: This study was a retrospective chart review with missing data for some variables. Future studies with a prospective design would increase quality of evidence. CONCLUSIONS: The majority of individuals prescribed an in-clinic exercise-based rehabilitation program are non-adherent. Patient diagnosis was the most important predictor of adherence. For those who were not adherent, important barriers include personal issues, insufficient insurance authorization and lack of geographic accessibility.