We have measured peak plasma concentrations of lignocaine and bupivacaine after dual injection peribulbar block and investigated the influence of adrenaline and hyaluronidase. Twenty-four patients were allocated randomly to one of four groups: (I) local anaesthetic alone (lignocaine 10 mg ml(-1)-bupivacaine 3.75 mg ml(-1)); (II) local anaesthetic with adrenaline (5 mu g ml(-1)); (III) local anaesthetic with hyaluronidase (75 iu ml(-1)); or (IV) local anaesthetic with adrenaline and hyaluronidase. Venous plasma concentrations of lignocaine and bupivacaine were measured in 24 patients using gas liquid chromatography before and at 5, 10, 15, 20, 25, 30, 45, 60, 90, 120, 180, 300 and 540 min after completion of the peribulbar injections. Main outcome measures were analysed using two-way analysis of variance. All patients, with one exception, received 10 ml of the local anaesthetic mixture. Overall peak plasma concentrations varied from 230 to 1910 mu g ml(-1) for lignocaine and from 160 to 1090 mu g ml(-1) for bupivacaine. Adrenaline significantly reduced peak plasma concentrations of lignocaine to 57% (P = 0.001) and bupivacaine to 61% (P = 0.004) compared with the non-adrenaline groups. Hyaluronidase had no significant effect on peak plasma concentrations of bupivacaine, which lignocaine and were 90% (P = 0.34) and 100% (P = 0.84) of the non-hyaluronidase groups. The area under the plasma concentration-time curves to 300 min (AUC(300)) behaved similarly. There was a reduction in AUC(300) for lignocaine (P = 0.005) and bupivacaine (P = 0.011) in the adrenaline groups compared with the non-adrenaline groups, in contrast with no significant effects of hyaluronidase on AUC(300) for lignocaine (P = 0.14) or bupivacaine (P = 0.53) compared with the non-hyaluronidase groups.