Background and purpose Chronic pain is a multifactorial problem. It is therefore important to assess the chronic pain patient from a variety of perspectives. A reduction in pain behaviour is a specifically cited aim of some clinical interventions. Pain behaviour measures have involved video recordings and subsequent ratings, or observations over a prolonged period of time. These methods have been criticised as time consuming and impractical in the clinical setting. The present pilot study developed a pain behaviour measurement tool, which could be delivered in 'real time' during a standardised functional assessment to give immediate feedback to clinicians and that could be used as an outcome measure. Methods Frequently occurring pain behaviours were identified, criteria were standardised, and a measurement tool and scoring system were developed. Forty-seven subjects with chronic pain were included in the inter-rater reliability study. Each subject was observed by 2 raters as they carried out a standardised physiotherapy functional assessment before and after a pain management programme. The occurrence of pain behaviour was scored using the developed measurement tool. Inter-rater reliability and internal consistency were assessed. Results and conclusion There was a high level of inter-rater reliability. 84 percent of the Kappa scores obtained fell between 0.61 and 1.0. Internal consistency was excellent (Alpha score 0.85 pre-programme and 0.84 post-programme). Given the limitations of self-report and self-monitoring methods of pain, pain behaviour and disability, a tool that more reliably captures this additional dimension of pain is appealing, and would be a valuable adjunct in the clinical assessment of the chronic pain patient