A number of pharmacologic and nonpharmacologic technologies are in current use to minimize perioperative homologous blood use. Clinical trials, many of them randomized controlled trials, have been done evaluating these approaches and have demonstrated their efficacy. However, data on safety has relied mostly on case reports, uncontrolled studies, and, for the pharmacologic agents, extrapolation from the nonsurgical setting. In this review I analyze the data from the randomized trials and the lower-level evidence studies to provide the best estimates in safety with these alternatives. In general, these alternatives are safe with proper dosing and monitoring of effects. With aprotinin, the primary concern is anaphylaxis, and this predominantly with re-exposure. With aprotinin and with the anti-fibrinolytics, increased venous thromboembolic risk has not been a consistent finding. Tranexamic acid use intraoperatively is advantageous, but postoperative use appears to have no advantage and may be associated with renal dysfunction. DDAVP is low-risk, provided it is not overused, which can induce hyponatremia. Autologous predonation probably has similar risks as homologous blood with respect to transfusion errors and bacterial infection. As with most medical interventions, we must be vigilant to prevent human error.