A 6-year follow-up study of a cross-section of the adult Danish population, based on data from the Danish Health and Morbidity Surveys in 1994 and 2000 is presented. The pain populations were identified through the pain intensity verbal rating scale (VRS) included in the Short Form 36. The 2000 survey also included a question on duration of pain (>6 months). Using this as the 'gold standard', a validation study was performed, which identified the highest accuracy (85%) at the VRS cut-off level: no pain, very mild, or mild pain (control group) versus moderate, severe, or very severe pain (pain group). The cohort comprised 2649 individuals, representative of the Danish population. Prevalence rates of pain in 1994 and 2000 were 13.5 and 15.7%, respectively. The cumulated 6-year incidence of pain development was 10.7%, and the cumulated incidence of pain recovery was 52.1%. During the investigated period, 9.2% of individuals moved from a 'no pain status' to a 'pain status', 7% moved from a 'pain status' to a 'no pain status', and 6.5% maintained their 'pain status'. Significant risk factors for pain development were female gender [odds ratio (OR) 1.5, 95% confidence interval (CI) 1.2-2.0], short education (OR 1.5, CI 1.0-2.2), poor self-rated health (OR 3.3, CI 2.4-4.7), and having at least one long-standing disease (OR 2.6, CI 2.0-3.4). Significant predictors for pain recovery were male gender, younger age, cohabitation status, good self-rated health, good mental health, having no long-standing disease, and having an annual income above 15,400 US dollars.