To investigate the sensitivity of polymerase chain reaction (PCR)-based detection of B-cell monoclonality and the effects of several variables, we analyzed 119 cases of B-cell lymphomas with proven IgH gene rearrangements, testing fresh-frozen and formalin-fixed materials in parallel. Using fresh-frozen tissue, 83 cases (70%) were positive with one-step PCR and high-stringency annealing. Two groups of false-negative cases were identified, Group I (16 cases) showing no PCR products, and Group II (20 cases) showing polyclonal smear patterns. Seminested PCR and/or lowered annealing stringency revealed nine additional positive cases in Group I but none in Group II. Amplification with bcl-2/JH primers resulted in five more positive cases. The overall positive rate in the fresh-frozen category was 81% (97 of 119). Parallel analysis was performed on formalin-fixed, paraffin-embedded material from 61 cases, and a high concordance rate (89%) was observed. The results indicated that fresh and formalin-fixed specimens are comparable in this PCR-based assay, and that the two groups of false-negative results can be accounted for by different reasons. In addition, we reviewed the current literature, discussed the diagnostic applications of this technique, and listed the core elements of a proposed PCR protocol that should be suitable for most laboratories.