The present study draws upon resource-based models of attention in suggesting that the processing of chronic and persistent pain is a task that demands the application of central and executive attention. If a chronic and persistent pain stimulus is demanding of central, attentional resources, it follows that it will compete with a second attention-demanding task for those limited resources. Here it is hypothesized that performance of an attention-demanding interference task will be detrimentally affected by the demands of persistent pain. In Expt 1, patients in high pain, patients in low pain and control subjects without pain performed an attention-demanding numerical interference task. There were no significant differences between any of the groups on any measure of performance. Expt 2 repeated Expt 1 with a more difficult and more complex task. Only when the task was at its most difficult and its most complex (i.e. at the greatest demand of limited resources) did those patients in high levels of pain (i.e. at the greatest demand of limited resources) show performance decrements. The results of both experiments are discussed in relation to the debate concerning the use of cognitive methods for pain control and in relation to the application of cognitive psychology to the study of chronic pain.