The Paediatric Pain Profile (PPP) is a 20-item behaviour rating scale designed to assess pain in children with severe neurological disability. We assessed the validity and reliability of the scale in 140 children (76 females, mean age 9 years 11 months, SD 4 years 7 months; range 1 to 18 years), unable to communicate through speech or augmentative communication. Parents used the PPP to rate retrospectively their child's behaviour when 'at their best' and when in pain. To assess interrater reliability, two raters concurrently observed and individually rated each child's behaviour. To assess construct validity and responsiveness of the scale, behaviour of 41 children was rated before and for four hours after administration of an 'as required' analgesic. Behaviour of 30 children was rated before surgery and for five days after. Children had significantly higher scores when reported to have pain than 'at their best' and scores increased in line with global evaluations of pain. Internal consistency ranged from 0.75 to 0.89 (Cronbach's alpha) and interrater reliability from 0.74 to 0.89 (intraclass correlation). Sensitivity (1.00) and specificity (0.91) were optimized at a cut-off of 14/60. PPP score was significantly greater before administration of the analgesic than after (paired-sample t-tests, p<0.001). Though there was no significant difference in mean pre- and postoperative scores, highest PPP score occurred in the first 24 hours after surgery in 14 (47%) children. Results suggest that the PPP is reliable and valid and has potential for use both clinically and in intervention research.