Purpose: Exercise therapy such as motor control training (MCT) has been shown to reduce pain and disability in people with low back pain (LBP). It is unknown which patients are most likely to benefit. This longitudinal cohort study aimed to: (1) retrospectively examine records from a large cohort of patients who received MCT treatment, (2) identify potentially important predictors of response to MCT and (3) test the predictors on an independent (split) sample derived from the original cohort of patients, using one group to identify the predictors and the other to test them. Methods: The response of 775 patients with LBP to MCT was classified as ‘improved’ or ‘not improved’ based on self-reported change in pain levels and symptoms. Measures were examined for associations with improvement and entered into a logistic regression model to classify patients as low, medium or high benefits of improvement with MCT. The model was tested on an independent sample. Results: A positive response was seen in patients with: no evidence of scoliosis [OR  = 4.0, 95% CI (1.7, 9.6)], LBP without associated groin pain [OR = 2.2, 95% CI (1.0, 5.0)], LBP which was chronic recurrent in nature [OR = 3.1, 95% CI (1.8, 5.3)] and poor results on muscle testing of the multifidus muscle [OR = 2.0, 95% CI (1.1, 3.7)]. Testing on an independent sample confirmed that patients could be classified into benefit groups. Conclusions: This study provides a first step towards assisting clinicians to select patients most likely to respond to MCT.