This study investigated the effects of vehicles on penetration and retention of lidocaine applied to sheep skin in vitro. Thoracic skin from two sheep was clipped of wool and stored at -20 °C, until used. Skin samples were defrosted and mounted in Franz-type diffusion cells, and then one of the following formulations, each saturated with lidocaine, was added: sodium lauryl sulphate (SLS) 0.5% in water, SLS 1% in water, dimethyl sulphoxide (DMSO) 50% in water (wt/wt), DMSO 100%, isopropyl myristate 100% (IPM), water alone, diethylene glycol monoethyl ether (DGME) 50% in water (wt/wt) and DGME 100%. The penetration of lidocaine in each skin sample was measured over 8 h. Significantly greater lidocaine skin concentrations and flux (J(SS)) were achieved with the nonaqueous vehicles, DMSO 100% (P < 0.00001 and P < 0.01, respectively), followed by DGME 100% and IPM (P < 0.00001 and P < 0.01, respectively). The lag time (t(lag)) for lidocaine penetration in the DMSO 100% vehicle was significantly shorter (P < 0.01) compared with all other vehicles except water. Improved transdermal penetration of lidocaine in the DMSO 100% vehicle was likely due to skin barrier disruption, as determined by differences in pre- and post-treatment transepidermal water loss (TEWL). This study has shown that nonaqueous vehicles enhanced penetration of lidocaine in sheep skin to a greater extent than aqueous vehicles, which has implications for topically applied local anaesthesia in sheep.