The radioallergosorbent test (RAST) was positive in 52.5% of 200 sera representing 200 food hypersensitivities from 108 patients with a history of definite immediate-type reactions to foods. Corresponding prick test was performed for 170 of the sera. The latter test was positive in 70%, the RAST was positive in 52%, and iether prick test or RAST was positive in 74%. It is concluded that the RAST is positive less frequently than the prick test in the diagnosis of immediate-type food allergy in clinically sensitive patients, but that the performance of both tests increases slightly the possibility of confirming the diagnosis. However, the RAST is useful for further evaluating positive prick tests with foods that do not correlate with clinical hypersensitivity.