The recently developed strategy for the detection of monoclonal B-cell populations, based on the selective amplification of predominant immunoglobulin heavy chain gene (IgH) rearrangement, is seen to be a suitable alternative to Southern blot analysis. The new technique uses a pair of consensus primers for variable (VH) and joining (JH) regions. We first tested the accuracy of this new approach on a broad series of 67 samples that had been well characterized by both Southern blot and immunohistochemical techniques before being subjected to blind testing. Our results show that this technique gave 100% specificity (absence of false-positive results) and approximately 70% sensitivity (30% false-negative results). The only exception was the presence of an IgH polymerase chain reaction ambiguous result in a case of Sezary's syndrome. The polymerase chain reaction technique was then applied to a panel of 27 frozen gastric endoscopy biopsy specimens following previous clinical suspicion of lymphoma. Monoclonality was detected in nine of 13 samples previously diagnosed as lymphomas and in one of two carcinomas. Further examination of the gastrectomy specimen in the latter case disclosed a B-cell lymphoma associated with the carcinoma. In spite of its limited sensitivity, the high specificity attained by this technique in the detection of monoclonality makes it a useful adjunct to routine morphologic criteria, as it is sometimes capable of detecting true positive cases that conventional morphology studies show to be negative.