This is the first controlled study of the frequency of back pain in a European caucasian population with diffuse idiopathic skeletal hyperostosis (DISH). Elderly patients admitted to hospital for reasons other than back pain were assessed for the presence of spinal DISH using the routine lateral chest radiograph films. A total of 106 probands (82 males, 24 females) with a mean age of 70 years fulfilled the criteria for DISH as defined previously. One hundred and seventy-eight patients (117 males, 61 females) not meeting these criteria were used as controls. The prevalence of back pain was assessed by a blinded interviewer using a structured questionnaire. Our primary hypothesis was that spinal DISH positive probands had not had back pain more often than controls. This controlled study showed no statistically significant difference in pain frequency between spinal DISH positive probands and controls at any spinal level. We conclude that back pain does not occur more often in radiographically defined DISH positive probands than in controls. The radiological finding of spinal DISH, as far as it does not lead to stenosis of the spinal canal or dysphagia, thus seems to be a finding without clinical relevance.