The integrase enzyme encoded by the human immunodeficiency virus plays an integral role in the viral life cycle, but is as yet unexploited as a clinical drug target. Integrase processes the viral DNA in the cytoplasm, translocates to the nucleus, and catalyzes viral DNA insertion into the host genome. A wide variety of chemical structures inhibit integrase in vitro, yet few of these apparently promising compounds have demonstrated similar efficacy in vivo. Multiple binding targets have been identified for different integrase inhibitors. These targets include the integrase enzyme prior to substrate binding, the viral DNA substrate, and the preintegration complex consisting of oligomeric integrase and the viral DNA. Some known inhibitors are effective only in the presence of divalent manganese as the active site metal ion cofactor, whereas others do not discriminate between manganese and magnesium ions. Integrase inhibition in response to ligand binding at one of multiple sites renders derivation of a simple set of structure activity relationships challenging. Progress toward this goal is reviewed in the context of experimental and theoretical structural information about integrase.