BACKGROUND Autotransfusion in pediatric anesthesia is beneficial in several clinical settings; however, more frequent usage is deterred by process-related costs and the fact that the absolute volume of blood returned may be minimal. OBJECTIVE Two autotransfusion devices (CATS, Fresenius Kabi AG, and Sorin Xtra, Sorin GmbH) with different technologies were evaluated to determine the minimum volume necessary to efficiently process salvage blood. METHODS Banked blood was diluted to simulate different clinical scenarios. Volume and red cell mass in returned blood were analyzed as a function of initial volume and red cell mass in the collection reservoir using incremental increases in the volume of blood added to the reservoir. RESULTS The volume of the returned processed blood for infusion was dependent on degree of dilution and blood volume in the reservoir. The CATS required a smaller initial minimum volume to process and successfully return; however, the Xtra provided a more efficient recovery at low initial collected volumes. Using test blood of various degrees of dilution, the CATS and Xtra were able to return blood for retransfusion after a mean initial red cell mass of 48.7 ml (±4.9 ml) and 53.7 ml (±5.0 ml) were processed, respectively. CONCLUSIONS Red cell mass rather than salvaged blood volume represents a reliable predictor of the successful use of an autotransfusion device. Measuring the hematocrit of the salvaged blood could improve the use of the devices. The investigated devices are likely to be roughly equivalent in effectiveness.