Twenty chronic patients, group 1, entered a hospital factory unit for one month and received varying increases of pay. They were compared with three control groups, each of ten patients. Thus group 2 received increases of pay similar to those of the group entering the factory, but remained outside the factory unit. Another, group 3, had already been in the unit for several months, and were discharged from the unit into other hospital employment for one month without change of pay. Another, group 4, remained in the factory throughout with no change of pay. Forty per cent. of each group was female. The control groups were matched with group 1 for age, duration of stay and group test results. Ranges and averages were: age, gp. 1, 36–66, 52·1, gp. 2, 34–65, 50·1, gp. 3, 38–68, 50·5, gp. 4, 21–63, 46·1; duration of stay, gp. 1, 1–35, 17·6, gp. 2, 3–27, 14·0, gp. 3, 2–32, 12·6, gp. 4, 2–27, 13·9 years. The distribution of patients between hospital wards was the same in each group. Eighty per cent. of each group had a diagnosis of chronic schizophrenia. Eight patients from group 1 and four from each of the control groups took part in an additional group discussion with one of the experimenters during the month trial period. A list of different topics were worked through, e.g. driving, smoking, taking risks, etc., and photographs from the 1963 Year Book and coloured advertisements were used, preferences and associations being asked for with each topic.