The relative potency of 12 commercial Alternaria extracts was analyzed by end-point skin test titrations and compared to in vitro measurements of potency, including (1) the radioallergosorbent test (RAST) inhibition procedure; (2) the direct RAST procedure; and (3) the protein nitrogen unit (PNU) content. Potencies determined by skin testing 10 sensitive patients were strongly correlated among the various patients. Measurements of potency by both RAST inhibition and direct RAST assay were strongly correlated to potency as measured by skin testing. In contrast, neither the weight:volume nor the PNU content bore any relationship to allergenic potency as measured by skin testing or by either of the RAST procedures. Extracts differed by as much as 3,000-fold in allergen content by skin testing. Moreover, the extracts appeared to contain different allergenic determinants when tested by RAST inhibition. RAST inhibition offered several technical advantages over the direct RAST procedure, in that only one solid-phase RAST reagent was required, slopes of dose-response curves could be more easily compared, and a greater discrimination in allergenic potencies among extracts could be made. The RAST appears to offer an excellent method for measuring the potency of allergy extracts, pending the isolation and characterization of actual allergens.