Detection of B-cell monoclonality using the polymerase chain reaction (PCR) promises the quick and cost-effective separation of monoclonal from polyclonal B-cell disease. However, the efficiency of the method has yet to be fully assessed, particularly with regard to disease type and selection of PCR primers. We have evaluated two approaches based on amplification of the immunoglobulin heavy chain gene using framework 2 (Fr2) and framework 3 (Fr3) region primers. Frozen tissue samples from 94 cases of low-grade B-cell lymphoma were investigated, all of which had previously been shown to be monoclonal by Southern blot analysis. Using a Fr2 primer, we were able to show monoclonality in 85 per cent of cases; with Fr3, 80 per cent of cases; and using both techniques in separate reactions, 90 per cent of cases. Thus, a significant false-negative rate exists with either primer which can be reduced by using both. We also found a difference in the efficiency of detection in different types of lymphoma; only 87 per cent of mucosa-associated lymphomas and centroblastic/centrocytic lymphomas were shown to be monoclonal, whereas all of the other lymphoma types tested were positive using one or both methods. We conclude that PCR detection of B-cell monoclonality allows rapid analysis of tissue samples, including paraffin-processed material. False-negative results which occur in some types of lymphoma can be reduced by the use of two or more primer combinations.