During the period April 1985 to March 1986, 217 blood donors were found to have moderate (syncopal) to severe (convulsive) reactions. This population was compared to 5630 randomly selected donors who did not have reactions. An examination of demographic, physical, and societal/emotional factors was conducted to determine if any were predictive of reactions in donors. The results of the research supported the hypothesis that first-time donors have a higher frequency of reactions (1.7%) than do repeat donors (0.19%). A review of the above predictive factors documented that, with regard to demographic factors, 1) the number of prior donations was inversely proportional to the risk of reaction; 2) the gender of the donor was not predictive; and 3) youth was a predictor of reactions. An analysis of the physical factors revealed that donors who reacted were of lower weight (mean, 153.7 lb) than those who did not (mean, 166.4 lb) and that systolic blood pressure was slightly lower in the group with reactions. Although the difference was significant (3 torr), it was not thought to be significant clinically. In a comparison of a group with systolic blood pressure ranging from 80 to 100 torr and a group with systolic blood pressure ranging from 120 to 140 torr, the first group had a 70-percent higher risk of reaction. Finally, with regard to the last category of societal or emotional factors, the research demonstrates 1) that the ingestion of caffeinated beverages was associated with a reduced risk of reactions; 2) that the food intake of donors who reacred was significantly different from that of those who had no reaction, but this difference was not thought to be clinically significant; and 3) that the duration between registration and the onset of phlebotomy was directly predictive of reaction status. The research indicates that first-time donor status and several specific demographic, physical, and societal or emotional factors are predictors of donor reactions.