The announcement that orders for long-term electrocardiograms were the subject of study in a community hospital was followed by a decrease in ordering rate of 30% from the corresponding quarter in the previous year. In succeeding quarters the decrement from corresponding quarters in the prior year was 21%, 27%, and 6%. Thereafter the ordering rate began to increase at a rate of 75% per year. In another community hospital in a contiguous town in which no surveillance was undertaken, the ordering rate for long-term electrocardiogram grew persistently at the rate of 42% per year over the same 3-year interval. The cardiologists at the community hospital under study behaved no differently as a group than other physicians in the community. The data strongly suggest that announced surveillance had the effect of diminishing long-term electrocardiogram tests ordered by community physicians by at least 20% for a period that lasts up to 9 months.