There is generally good evidence that pain management interventions that include self-management strategies can substantially reduce disability and improve psychological well-being in patients with chronic pain. Reductions in unhelpful responses, especially catastrophising and fear-avoidance beliefs, have been established as key contributors to these gains. In contrast, there is surprisingly little evidence that adherence to self-management strategies contributes to achieving these outcomes. Difficulties in defining and measuring the use of pain self-management strategies have been obstacles for this research. Using a pragmatic way of assessing the practice of specific strategies this study investigated their ability to account for changes in pain, disability and depressive symptoms after a 3-week cognitive-behavioural pain management program. The post-treatment outcomes on these dimensions were found to be statistically and, for many, clinically significant. Consistent with previous research, reductions in catastrophising and fear-avoidance beliefs, and increased pain self-efficacy beliefs, were also associated with these gains. But the key new finding was that there was a clear gradient between adherence to specific self-management strategies and reductions in pain, disability and depressive symptoms. Furthermore, adherence to the self-management strategies was predictive of better outcomes even after controlling for the moderating effects of initial catastrophising, fear-avoidance and pain self-efficacy beliefs.