"The Göteborg quality of life instrument" (GQL-instrument) has been used to assess the quality of life of men born in 1913 and 1923. On a population basis, it was possible to show that the well-being variables were stable over time and that excellent well-being showed a great variation. A high level of well-being was common in variables such as family and housing but more uncommon in variables such as fitness, vision, hearing and memory. Symptoms were often significantly related to biomedical variables such as body mass index, blood pressure, lung function, blood lipids, fasting blood sugar and fasting insulin. The GQL-instrument seems to provide a reliable and stable assessment of well-being and symptoms and is useful both as a descriptive tool, and as a help in evaluating treatment, and it also has predictive power.