Abstract Patients with chronic low back pain present with a mixture of symptoms and signs. Some are a direct consequence of physical pathology whereas others are attributable to associated and appropriate psychological and behavioural changes. At times the latter may be out of keeping with the degree of physical pathology and thus have specific significance in terms of the affective and cognitive disturbances that are also present and which may be the basis for abnormal illness behaviour. In an attempt to demonstrate more clearly the relationship between physical, psychological and behavioural components of illness, this paper draws on two data sets in patients with low back pain. The first explores the relationship between behavioural symptoms and signs, objective physical impairment, pain and disability and psychometric measures of distress together with scales making up the illness behaviour questionnaire (IBQ) of Pilowsky and Spence. A second data set is used to assess the value of the IBQ in understanding how psychological distress and behavioural signs and symptoms are related to the outcome of surgical treatment. The results gained reveal that behavioural symptoms and signs are directly related to the physical severity of low back disorder, the patient's report of pain and disability and the outcome of surgical treatment. Scores on the IBQ were strongly related to measures of affective disturbance and psychological distress. More specifically the disease affirmation scale of the IBQ, incorporating scales for disease conviction and psychological versus somatic focussing was an important dimension in relation to the behavioural symptoms and signs, thereby confirming results gained by other workers. Disease conviction and lack of response to clinicians' reassurances regarding illness — a situation in which abnormal illness behaviour is often deemed to exist — should not be seen simply as a function of the disease process, but more as a psychological coping mechanism for certain individuals under stress. The significance of this observation is discussed in relation to decisions regarding the overall assessment of chronic pain patients and their treatment.