This paper employs classical concepts of diminishing marginal utility to demonstrate that risk-aversion can increase the perceived value of diagnostic procedures and thus raise optimum diagnostic expenditures. The theory is applied to a model in the spirit of Phelps and Mushlin's initial technology assessments. The specific evaluation is the cost-effectiveness of somatostatin receptor scintigraphy used to detect distant metastases of carcinoid liver tumours in a patient otherwise eligible for surgical resection of the liver. Data for the model are taken from published sources and financial databases, when available, and otherwise from a senior clinician's experience (LKK). The quantitative results indicate that receptor scintigraphy may have two beneficial impacts to risk-neutral individuals. First, it may reduce the combined costs of therapy and treatment because the diagnostic procedure costs less than the expected savings generated by avoiding inappropriate surgeries. Second, it may improve the patient's expected health-status-adjusted life years (HSALY) because the information allows physicians to better match treatment to the cancer's stage. Finally the paper demonstrates that risk aversion, as embodied in classical diminishing marginal utility applied to health status, can increase the value of the diagnostic tests and can lead the patient to choose a less beneficial treatment. An illustrative risk-averse utility function changed the optimum treatment from surgery to chemotherapy and increased scintigraphy's benefit by 500%.