Early validation of novel targets for the treatment of pain remains a key challenge, despite the recent technological advances which aid more rapid generation of pharmacological tools and molecular techniques that allow in vivo manipulation of gene transcription. Use of genetically modified animals, particularly 'knockouts', has been critical in evaluating the function of targets for which pharmacological manipulation has proved difficult. This review discusses recent findings from such investigations for three of the larger target classes, and highlights the relative merits and drawbacks of this methodology compared with other methods for the functional investigation of the role of individual molecular targets in pain.