Novel antiretroviral drugs offer different degrees of improvement in activity against drug-resistant HIV, short- and long-term tolerability, and dosing convenience compared with earlier drugs. Those drugs approved more recently and commonly used in treatment-experienced patients include the entry inhibitor enfuvirtide, protease inhibitors (PIs) [darunavir and tipranavir], a C-C chemokine receptor (CCR) type 5 antagonist (maraviroc), an integrase inhibitor (raltegravir) and etravirine, a non-nucleoside reverse transcriptase inhibitor (NNRTI). Novel agents in earlier stages of development include a CCR5 monoclonal antibody (PRO 140) administered subcutaneously once weekly, once-daily integrase inhibitors (elvitegravir and S/GSK1349572), and several nucleoside (nucleotide) reverse transcriptase inhibitors and NNRTIs. Bevirimat, a maturation inhibitor, has compromised activity in the presence of relatively common Gag polymorphisms. Viral suppression is necessary to control the evolution of drug resistance, reduce chronic immune activation that probably underlies the excess morbidity and mortality in HIV-infected patients, and reduce viral transmission, including transmitted drug resistance. In general, the proportion of viraemic patients who achieve suppression increases with the number of active pharmacokinetically compatible antiretroviral drugs in the regimen. In the ANRS139-TRIO trial, 86% of highly treatment-experienced patients treated with darunavir-ritonavir, etravirine and raltegravir had HIV RNA <50 copies/mL at 48 weeks. In patients who had received at least 12 weeks of a stable regimen and had no darunavir resistance-associated mutations, once-daily darunavir boosted with ritonavir 100 mg was virologically noninferior with better lipid effects than with the twice-daily dosing, which requires a 200 mg total daily dose of ritonavir. Raltegravir plus a boosted PI is being investigated for second-line therapy in patients not responding to NNRTI-based first-line treatment in resource-limited settings (RLS). However, concerns about this potential strategy include the low barrier against resistance of raltegravir, limited penetration of some PIs into the CNS and the unknown impact of integrase polymorphisms seen more commonly in non-B subtype HIV-1. In patients who have already achieved viral suppression, novel agents may be used to simplify the dosing schedule, lower costs (such as by switching to boosted PI monotherapy), reduce adverse events or preserve antiretroviral drug options, especially since the absence of an HIV eradication strategy implies the need for life-long combination antiretroviral therapy. Switching enfuvirtide to raltegravir eliminated painful injection-site reactions without compromising virological suppression. Two studies found different virological outcomes when patients were switched from lopinavir/ritonavir to raltegravir, but there was an improvement in the lipid profile. Simplifying to darunavir-ritonavir monotherapy after suppression of plasma HIV RNA to <50 copies/mL has been found to be safe with no emergence of resistance in cases of viral rebound, but longer-term data are needed. The initial suggestion that maraviroc may possess unique CD4+ T-cell boosting effects was not confirmed in several clinical trials. Improved understanding of HIV pathogenesis has opened new frontiers for research such as identifying the sources, consequences and optimal management of residual viraemia in those with plasma HIV RNA <50 copies/mL. Globally, however, one of the most urgent priorities is providing the increasing number of treatment-experienced virologically failing patients in RLS with access to optimal treatment, including those treatments based on novel antiretroviral agents.