The Coping Strategies Questionnaire (CSQ) (Rosenstiel and Keefe 1983) is the most widely used measure of pain coping strategies. To date, with one exception (Tuttle et al. 1991), studies examining the factor structure of the CSQ have used the composite scores of its 8 a-priori theoretically derived scales rather than the 48 individual items. An examination of the match between the 8 theoretically derived scales and scales empirically extracted from an item analysis is lacking. Accordingly, the CSQ was administered to 126 chronic pain (whiplash) patients. Factor analyses of the individual items revealed an 8-factor structure to be uninterpretable. Of the 2-9-factor solutions tested, the 5-factor structure was the most interpretable: Factor 1, Distraction; Factor 2, Ignoring Pain Sensations; Factor 3, Reinterpreting Pain Sensations; Factor 4, Catastrophizing; Factor 5, Praying and Hoping. Eighteen Ph.D. or M.D. level clinicians classified items into their corresponding category with a high degree of accuracy (on average, 90.2%), attesting to the face and construct validity of the subscales. Four subscales, Catastrophizing, Reinterpreting Pain Sensations, Praying and Hoping and (to a lesser degree) Ignoring Pain Sensations, correspond with parallel subscales proposed by Rosensteil and Keefe (1983). The fifth subscale, Distraction, is comprised of items from their Diverting Attention and Increasing Activity Level subscales, suggesting that cognitive and behavioural distraction comprise 1 rather than 2 coping strategies. That CSQ items on the original Coping Self-Statements and the Increasing Pain Behaviour subscales failed to load consistently on any factor suggests that they do not reliably measure distinct coping strategies.(ABSTRACT TRUNCATED AT 250 WORDS)