AIMS: To evaluate the specificity of standard and fluorescence based (Genescan) polymerase chain reaction (PCR) immunoglobulin heavy chain (IgH) gene rearrangement analysis in complete and microdissected paraffin wax embedded sections from lymphoid proliferations. METHODS: PCR IgH gene rearrangement analysis of whole sections and microdissected fragments (n = 62) from paraffin wax embedded reactive lymph nodes (n = 6) and tonsils (n = 3). Amplificant analysis used both standard methods and automated high resolution fluorescence based quantification and size determination using GENESCAN software. RESULTS: Whole tissue sections were consistently polyclonal in control experiments. IgH gene amplification was successful in 59 of 62 microdissected fragments; only two of 59 showed a polyclonal rearrangement pattern, the remainder being oligoclonal or monoclonal. Reanalysis was possible in 33 samples; six showed reproducible bands on gel analysis and satisfied accepted criteria for monoclonality. Use of high resolution gels with Genescan analysis improved sensitivity and band definition; however, three samples still appeared to be monoclonal. CONCLUSIONS: These results confirm that PCR based IgH gene rearrangement analysis is a sensitive and specific method for demonstrating B cell clonality in whole paraffin wax embedded sections. However, oligoclonal and monoclonal rearrangement patterns are regularly encountered in small tissue fragments from otherwise unremarkable reactive lymphoproliferations, possibly because of preferential priming or detection of local B cell clones. Data from clonal analysis of small, microdissected or lymphocyte poor samples must be evaluated critically. It is recommended that analyses should be run in parallel on at least two tissue specimens. Only reproducible bands present in more than one sample should be considered to be suggestive of neoplasia.