As part of a retrospective study into the prevalence of the t(14;18) translocation in B-cell lymphomas, we assessed the suitability of the polymerase chain reaction (PCR) to amplify the t(14;18) major breakpoint region (MBR) in frozen and formalin-fixed tissue. Considering Southern blotting as a standard, the sensitivity of PCR was 81%. Of the various procedures used to extract DNA from paraffin-embedded tissue (PET), proteinase K digestion in the presence of nonionic detergents gave the highest yield and quality of DNA and the most efficient amplification rate. Using this method, excellent amplification rates (100%) were obtained for both the beta-globin control sequence and the MBR t(14;18) for fixed follicular lymphoma specimens collected in the previous 2 to 6 years (n = 27). Of nine older PETs, PCR on six gave inconsistent results, probably because of the poorer-quality substrate used for amplification. Specimens exposed to formol sublimate or formalin-acetic acid-alcohol were as suitable for amplification as tissues fixed in neutral-buffered formalin. The overall incidence of the MBR t(14;18) in all follicular lymphoma specimens as detected by both Southern blotting and PCR was 59% (23 of 39).