The incidence and clinical characteristics of anaphylactoid reactions to intravenous asparaginase were assessed in 196 patients given E. coli asparaginase and 49 patients given Erwinia asparaginase. All patients were given a 50 IU intravenous test dose followed in 30 min by the full dosage (10,000 IU/m2), if no reaction occurred to the test dose. Twenty-nine of 196 patients (14.8%) given E. coli asparaginase had an anaphylactoid reaction, occurring after their first through 12th doses. The probability of an anaphylactoid reaction was significantly greater in those patients not receiving concomitant prednisone-vincristine and patients with a hiatus between courses of asparaginase therapy. By logistic regression analysis, other variables such as age, sex, race, diagnosis, total number of doses and concurrent methotrexate or arabinosylcytosine did not contribute significantly to the probability of a reaction. Twenty-three of the patients who had reacted to E. coli asparaginase and 26 patients who had not reacted to E. coli asparaginase were subsequently given Erwinia asparaginase. Seven of these 49 patients (14%) had an anaphylactoid reaction. The probability of a reaction to Erwinia asparaginase was significantly related to a prior reaction to E. coli asparaginase, concomitant prednisone-vincristine therapy, total number of asparaginase doses, number of prior E. coli asparaginase doses, and diagnosis, when assessed by a logistic regression model. However, after adjusting for prior reaction to E. coli asparaginase and the total number of asparaginase doses given, the other variables did not contribute significantly to the probability of a reaction. Only 5/29 patients reacting to E. coli asparaginase and 1/7 reacting to Erwinia asparaginase had a reaction to the test dose. None of the reactions were fatal.