Alternative splicing of mRNAs occurs in the majority of human genes, and most differential splicing results in different protein isoforms with possibly different functional properties. However, there are many reported splicing variations that may be quite rare, and not all combinatorially possible variants of a given gene are expressed at significant levels. Genes of interest to pharmacologists are frequently expressed at such low levels that they are not adequately represented in genome-wide studies of transcription. In single-gene studies, data are commonly available on the relative abundance and functional significance of individual alternatively spliced exons, but there are rarely data that quantitate the relative abundance of full-length transcripts and define which combinations of exons are significant. A number of criteria for judging the significance of splice variants and suggestions for their nomenclature are discussed.