Exercise is a core treatment for persistent non-specific low back pain (NSLBP), but results from randomised controlled trials (RCTs) of exercise typically show only small to moderate standardised mean differences (SMDs) compared to non-exercise controls. The choice of primary outcome, and relationship to the specific targets of exercise may influence this. This systematic review aimed to explore whether primary outcomes match the exercise treatment targets used in NSLBP RCTs and the potential impact of matching on SMDs. Included RCTs were conducted with patients with persistent NSLBP, compared exercise to no exercise, with sample sizes >60 per arm. Screening, data extraction and risk of bias assessment were independently undertaken by paired reviewers. Of 19272 initial titles, 27 RCTs were included with 31 treatment targets and 6 primary outcome domains identified. Only 25% of included RCTs had primary outcomes that matched the treatment targets. SMDs of exercise versus comparison arms were observed to be larger in the matched (SMD 0.54 (95% CI 0.23 to 0.85), p=0.0006) compared to the unmatched category (SMD 0.22 (95% CI 0.01, 0.44) p=0.04) but this difference was not statistically significant (p=0.10). These exploratory findings may have implications for future teams developing RCTs of exercise for NSLBP and warrant further investigation in larger datasets. Perspective This review was an exploratory study that investigated the primary outcome and treatment targets used in RCTs of exercise for NSLBP. The SMDs of the matched group were descriptively larger than those of the unmatched group, but further analysis with larger sample sizes is required to have confidence in these results.Copyright © 2020. Published by Elsevier Inc.