Long-term outcome in back pain is related mainly to cognitive factors such as pain-related beliefs and expectations. Most research has been performed on patient samples. This study aimed at investigating changes over time in reported back pain, pain intensity, disability, health care consumption, and sick leave as well as biopsychosocial factors over a 12-month period. A second aim was to identify predictors of reported pain, pain intensity, disability, health care consumption, and sick leave. As parts of a large back pain sample from a general population (n = 1,024), two groups-one with first-episode pain (n = 77) and one with long-term pain (n = 302)-responded twice to a self-administered questionnaire. Among participants reporting pain at both assessments, changes over time were analyzed and predictive models were tested. Generally, the results demonstrated overall stability in the self-reports over time. However, reported pain decreased in both groups, while pain catastrophizing and pain expectations increased in the first-episode group. Pain intensity and disability were predicted in regression models including four cognitive factors and initially reported levels of pain intensity and disability. The significance of pain-related beliefs and expectations both in early and later stages of a back pain condition is pointed out. The results in this study based on a sample from the general population are in line with previous research on patient samples.