Oligoclonal or clonal T-cell expansions, presumed to be antigen driven, are frequently sought and followed for diagnostic and prognostic purposes, as well as to understand more about their natural history. Techniques based on conservation of T-cell receptor CDR3 length are increasingly widely used, often without assessment of sensitivity or specificity. We present a comparative evaluation of a novel modified heteroduplex technique and a CDR3-length-based assay. Dilution of a known clone in a mixed T-cell population shows that in our hands the heteroduplex technique is at least 10-fold more sensitive than the CDR3-length-based assay. However, even with this level of sensitivity, we do not detect clonal expansions in unstimulated CD4+ T cells. This contrasts with the frequent detection of CD8+ clones in fresh samples and suggests different mechanisms of clonal homeostasis in the two subsets. We show that both techniques detect functional expansions after in vitro stimulation with a recall antigen. The distinct molecular footprint seen with the heteroduplex technique allows reproducible follow up of specific clonal expansions. We have exploited this to demonstrate that the repertoire of clones expanded by in vitro tetanus toxoid stimulation shows stability within an individual, implying long-term maintenance of multiple CD4+ clones.