There is a growing conviction that medical technologies are major contributors to escalating costs, and regulating them is generally viewed as the least contentious way to control expenses in the 1980's. Five forms of technology control are being discussed or developed. All aim to reduce costs by controlling big, expensive technologies in the class of computed tomographic (CT) scanning. We present evidence that technologies such as the CT scanner account for far less of the growth in medical expenditures than do the collective expenses of thousands of small tests and procedures. Furthermore, we suggest that each strategy for controlling large technology involves substantial practical and conceptual problems that would severely limit its effectiveness. We thus suggest a shift away from attempts to harness the big technologies, and toward incentives to encourage the more discerning use of all technologies. To this end, we propose changes in physician reimbursement and education and expanded insurance incentives to encourage physicians and hospitals to be more selective in the use of technology.