Background and objective: Sub-Tenon's block is usually delivered by the infero-nasal (IN) approach, but occasionally this may not be possible. The infero-temporal (IT) approach has been described, but data is not available on its efficacy. Methods: One hundred patients undergoing cataract extraction were randomized to receive an IN or IT sub-Tenon's injection of lidocaine 2% with hyaluronidase 15 IU mL−1. Akinesia was assessed using the Brahma scale at 0, 2, 4, 6 and 8 min. Injection, intraoperative and postoperative pain scores (verbal analogue score, 0–10) were noted, along with the incidence of sub-conjunctival haemorrhage and chemosis. Results: There were no differences in patient characteristics data, or mean volume of administered local anaesthetic solution (3.3 (SD = 0.4) mL). There were no significant differences between groups in terms of onset of akinesia. Mean akinesia scores at 2, 4, 6 and 8 min were 2.7, 1.1, 0.4 and 0.2 for Group IN, compared to 2.2, 0.9, 0.8 and 0.3 for Group IT. Chemosis occurred in 14 patients in Group IN, compared to 22 in Group IT (P = 0.21).A sub-conjunctival haemorrhage was noted in 14 patients in Group IN and 19 patients in Group IT (P = 0.52). No patients required supplementary injections. Mean pain scores for the injection, intraoperatively and postoperatively were 0.9, 0 and 0 for Group IN, compared to 1.1, 0 and 0 for group IT. The surgeons scored all the blocks as ‘good’ except for one patient in each group. Conclusions: The IT approach provides an equally rapid onset of block, without a significant increase in complications.