Histone deacetylase (HDAC) inhibitors can induce differentiation, cell cycle and growth arrest or in certain cases apoptosis in cancer cells. In a remarkably short period of time, especially considering that their mechanism of action remains largely undefined, HDAC inhibitors have realized both success and failure as therapeutics for cancer in clinical trials. Notably, the pleiotropic HDAC inhibitors, suberoylanilide hydroxamic acid (SAHA) and depsipeptide, have shown efficacy in a wide range of cancers, in particular for cutaneous T-cell lymphoma (CTCL), and are progressing in phase II clinical studies. However, evidence is accumulating that specific HDAC enzymes are important with respect to clinical efficacy, calling the usefulness of the classical inhibitors into question. Class I enzymes are being heralded as the most clinically relevant, however, this is still controversial and much of the information is in the private domain. Nevertheless, the potential to alter the expression of a more focused, disease-related subset of genes and to limit adverse effects has prompted the development of isoform-specific HDAC inhibitors. Here, we consider the growing view that broad-spectrum HDAC inhibitors may be superseded by more specific compounds.