A number of studies have commented on the inadvisability of transfusing unused blood given by autologous and directed donors to patients other than those for whom the blood was originally given (crossing over). It has been stated that these bloods are dangerous and should not be used for allogeneic transfusion. To test this hypothesis, data were taken from a pool of hospital donors. These show that if the donor groups of allogeneic, directed, and autologous are further divided into sets of first time and multiple donors, the incident of infectious disease markers is no higher in one set than the other except for the allogeneic first time donor. Therefore, the ''dangerous'' donor is actually more likely a first time donor. Following this reasoning, the use of all blood from first time donors should be suspect.