A monoclonal B cell population is the hallmark of B cell neoplasms including cutaneous B cell lymphomas (CBCLs). We modified and tested several polymerase chain reaction (PCR)-based assays involving amplification of immunoglobulin heavy chain (IgH) gene rearrangements to optimize assays specifically for cutaneous lymphoid infiltrates. We achieved greatest sensitivity with an assay employing IgH consensus primers complementary to the framework 3 portion of the upstream variable region and the downstream joining region. We studied 12 CBCLs, 6 nodal lymphomas and 7 cell lines. In 17/25 of these B cell neoplasms (84%), we detected one or two dominant bands, consistent with one or both IgH alleles being rearranged in the neoplastic B cell clone. As expected, IgH PCR assays produced diffuse smears in agarose gels or complex ladders in polyacrylamide gels when polyclonal B cell controls (blood and tonsil) were analyzed. However, in normal skin and non-CBCL skin lesions, one or a small number of discrete bands were sometimes detected. In certain cases, this made it difficult to distinguish true positives (monoclonal CBCL) from false positives (clonally restricted benign B cells). Correlation with immunophenotyping confirmed that false positive results were confined to samples with sparse or immunohistologically undetectable B cell infiltrates. Pseudoclonal bands showed variable sizes in repeat PCR reactions and could be distinguished from monoclonal bands by polyacrylamide gel electrophoresis of pooled triplicate PCR products. These findings suggest that molecular diagnosis using IgH PCR assays is best suited for B-cell-rich infiltrates, and can be problematic when applied to suspected T-cell-rich CBCLs, cutaneous T cell lymphomas, or other lesions containing only few B cells unless one is cognizant of the potential pitfalls. Furthermore, these results demonstrate the presence of rare B cells in normal skin and immunohistologically defined cutaneous T cell infiltrates. This correlates with recent reports of sparse B cells within the lymph draining from normal skin and may represent molecular evidence for a trafficking B cell component of the skin-associated lymphoid tissue (SALT). It also suggests a candidate B cell subset for the pathogenesis of cutaneous lymphoid hyperplasia and CBCLs.