Dextran reactive antibodies (DRA) were studied in 123 patients having experienced dextran-induced anaphylactoid reactions (DIAR) during 1970–1975. No evidence for reaginic DRA was obtained by radioallergosorbent technique and passive cutaneous anaphylaxis in Cynomolgus monkeys; total IgE levels were within normal range. It is concluded that DIAR are not mediated by dextran-specific reagins. Further, no reaginic antibodies against potential contaminants from the dextran manufacturing process were demonstrable. In two population samples of normal human sera from Sweden and Germany hemagglutinating DRA (IgG, IgA, and IgM classes) were found in 63 and 74%, high titres (16–256) comprising 14 and 25%. In dextran reactors a direct positive correlation between titres of hemagglutinating DRA and increasing severity of DIAR was observed. The accumulation of high DRA titres in severe reactions may be taken as circumstantial evidence for the causal role of hemagglutinating DRA in these cases. However, if high titres of DRA alone were responsible for triggering DIAR, the expected frequency would be more than thousand times higher than the reported global incidence of DIAR. To explain this discrepancy, involvement of certain Ig classes or subgroups, possibly in combination with other predisposing factors is suggested. In mild reactions DRA appear to play a negligible role. Positive dextran wheal and flare reactions, often correlated with high titres of hemagglutinating DRA, were seen in 32% of dextran reactors, indicating that skin tests are of limited predictive value. No significant difference between sexes, age groups, or pre- and intraoperatively started dextran infusions was observed; association with certain diseases was not apparent.