A university-based hospital consulting group reviewed six studies of Michigan hospitals retrospectively in 1975. The studies represented all those done between 1967 and 1971 requiring forecasts of acute bed supply and service needs. The original studies developed forecasts using empirical studies of patient origin and rigorously prepared authoritative forecasts of county populations. The 1975 review compared forecasts of population, service population, and bed need against current values and also interviewed clients to assess retrospective satisfaction with the recommendations. Although the consultants strove steadily to minimize the bed supply and base population forecasts were accurate, the studies overestimated bed needs. Further, the clients were often dissatisfied with the original recommendations, and frequently acted to exceed them. Comparing the 1975 actual with what would now be recommended by the consultant indicates that the “'error” cost the communities about $50 per person per year.