Drug approval numbers remain stagnant, the cost of bringing a drug to market is higher than ever and safety issues have come to the fore following the Vioxx debacle. So what can be done to improve the situation for pharmaceutical R&D? On the Analyst's Couch, Jones and Clifford discuss how early-stage alliances between biotech and pharma companies are increasingly being seen as part of the solution, and FitzGerald calls for the establishment of 'translational medicine and therapeutics' — an integrated discipline that could help prevent some of the recent problems encountered with drugs such as Vioxx. Also taking the withdrawal of Vioxx as an example, Carpenter and Ting provide a theoretical discussion of the political logic of regulatory error and the implications for future regulatory policies. Drug–drug interactions (DDIs) have been another common reason for withdrawal of approved drugs, and Wienkers and Heath highlight commonly encountered problems with predictions of DDIs from in vitro experiments. The limitations of animal models are also recognized as an important area for improvement, and as part of our model organism series Buras et al. critically review the models for sepsis. The final two reviews this month present novel therapeutic approaches. White et al. summarize strategies for targeting chemokines for the treatment of neuropathic pain, and Kwong and colleagues look at the potential of targeting helicases — the motor enzymes that unwind double-stranded nucleic acids — to meet challenges posed by viral infections. There is also a pressing need for new drugs to combat bacterial infections, and 'Fresh from the Pipeline' highlights tigecycline, the first member of a new family of antibiotics.