OBJECTIVE To compare a protocol of evidence-based conservative care with usual care for acute nonspecific low back pain (LBP) of less than 6 weeks' duration. DESIGN Parallel-group randomized trial. SETTING Three practices in the United Kingdom. PARTICIPANTS Convenience sample of 149 eligible patients were invited to participate in the study, with 118 volunteers being consented and randomly allocated to a treatment group. INTERVENTIONS The experimental group received evidence-based treatments for acute nonspecific LBP as prescribed in a structured protocol of care developed for this study. The control group received usual conservative care. Participants in both groups could receive up to 7 treatments over a 4-week period. MAIN OUTCOME MEASURES Oswestry Low Back Disability Index (ODI), visual analog scale (VAS), and Patient Satisfaction Questionnaire, alongside estimation of clinically meaningful outcomes. RESULTS Total dropout rate was 14% (n=16), with 13% of data missing. Missing data were replaced using a multiple imputation method. Participants in both groups received an average of 6 treatments. There was no statistically significant difference in disability (ODI) scores at the end of week 4 (P=.33), but there was for pain (VAS) scores (P<.001). Interestingly, there were statistically significant differences between the 2 groups for both disability and pain measures at the midpoint of the treatment period (P<.001). Patient satisfaction with care was equally high (85%) in both groups. Minimally clinically important differences in scores and number needed to treat scores (NNT<6) indicated that the experimental treatment (protocol of care) offered a clinically meaningful benefit over the control treatment (usual care), particularly at the midpoint of the treatment period. CONCLUSIONS Overall, the 2 treatment groups were similar based on primary or secondary outcome measure scores for the full treatment period (4 weeks, with up to 7 treatments). However, there were statistically significant and clinically meaningful differences in both disability and pain scores at week 2 (midpoint) with 4 treatments, suggesting that the protocol of care had a more rapid effect than usual care.