We thank Aveline et al.1 for their detailed account of central nervous system (CNS) toxicity following ultrasound-guided sciatic and saphenous nerve block using lidocaine and ropivacaine in a patient scheduled for hallux valgus surgery. They reported clinical signs compatible with systemic absorption of local anaesthetics and initiated treatment with 20% intralipid as 100 ml bolus doses 2 min apart. Unfortunately, this intervention was unsuccessful and general anaesthesia was required. The authors concluded that intralipid had been ineffective and suggested an interaction between local anaesthetic agents, intralipid, and carbamazepine treatment the patient was taking for trigeminal neuralgia, may have been a contributory factor. We agree that the patient exhibited local anaesthetic systemic toxicity (LAST) and wish to suggest possible explanations for the apparent failure of intralipid in this case. In their calculations of the local anaesthetic dose administered, the authors appear to have omitted the saphenous nerve component of their technique of 100 mg lidocaine, which, if included means the patient received a total dose of 400 mg or 7.02 mg kg−1 of plain lidocaine. The plasma lidocaine concentration of 2.3 μg ml−1, recorded 25 min after the block, is consistent with levels previously reported with injection of the same dose of lidocaine in the subcutaneous tissue of the abdominal wall.2 This would support the notion that LAST most likely represents systemic absorption from correctly placed peripheral nerve blocks rather than primarily from an intravascular injection. If the additive effect of the ropivacaine element of the block is taken into consideration in overall dose calculations,3 this may explain why CNS signs of LAST became apparent. Given the absence of significant cardiovascular, metabolic or electrolyte disturbance, all of which may have contributed to the severity of LAST, we agree that drug interaction may have been relevant in this case.4 All three drugs rely on protein binding of 75% or more for transport within the blood and all three are lipid soluble to varying degrees, with log P values for carbamazepine, lidocaine and ropivacaine of 2.73, 2.36 and 3.11, respectively. It is possible that competition for protein-binding sites between drugs resulted in a greater proportion of unbound local anaesthetics manifesting as systemic toxicity. Thus, for the lipid sink sequestration activity of intralipid to be fully effective, we consider this may have required administration of a proportionately larger volume of intralipid than usual. Previously intravenous lipid emulsion (ILE) has been reported to reverse LAST completely after presumed intravascular injection. We are aware of two published reports in which recovery from LAST occurred by early bolus administration of 20% intralipid followed by an infusion to a total volume of 500 ml.5,6 Although the time taken to restore a full level of consciousness did vary, in these cases the ILE reversed the LAST rapidly enough to avoid tracheal intubation. We advocate the early use of ILE but not as an alternative to safe local anaesthetic dosing and other appropriate resuscitation measures. Guidance for treatment of LAST has been posted on the Lipid Rescue website (http://www.LipidRescue.com) and elsewhere.7 For adults, these recommendations generally advocate an initial bolus dose of 100 ml (approximately 1.5 ml kg−1) 20% intralipid, followed by a rapid infusion of 400 ml over 20 min. Two further 100 ml boluses may be given and an infusion continued until signs of LAST have been reversed, if necessary continuing until a total volume of 12 ml kg−1 has been administered. We are curious as to why intravenous lipid treatment was suspended when only 200 ml had been given. Although we can only speculate at what might have happened had the volume of intralipid administered been bigger, this case is of value because it seems to confirm that the effects of ILE in the clinical setting are dose-dependent and perhaps the dose administered in this case was inadequate.