We read with interest the study by Gupta et al. on the use of magnesium as an adjuvant during general anaesthesia [1]. We would like to share some observations on the study. Magnesium efficacy and usefulness in different clinical settings has recently been extensively reviewed [2]. However, clinical use of a drug warrants examination of its safety profile and cost effectiveness. Currently, the use of magnesium in clinical practice is limited to emergency and life-threatening conditions [3] because its safety profile has not yet been extensively studied. In their study, Gupta and colleagues used 30 mg.kg−1 of magnesium sulphate prior to induction and 10 mg.kg−1.h−1 by continuous infusion during surgery, but they didn't explain the rationale for this dose regimen. Most previous studies have used doses between 40 and 60 mg.kg−1 at induction, followed by 8–15 mg.kg−1.h−1 intra-operatively [4-7]. They compared time to recovery of the twitch height to 25% (T3) and time to reach a BIS value of 80 following discontinuation of rocuronium (30 min before end of surgery) and propofol (at skin closure) infusions, respectively. However, they did not describe when the magnesium sulphate infusion was stopped. They concluded that there was a significant delay in recovery of T3 and recovery of the BIS to 80 in the magnesium group. A rise in serum magnesium from 1.99 to 2.17 mg.dl−1 was observed postoperatively, but this did not exceed the normal physiological range. Serum magnesium concentration was not measured intra-operatively. We believe that for an accurate interpretation of the results such intra-operative concentrations are important. As the serum magnesium concentration did not exceed the physiological range, and we don't know the intra-operative concentration, it is difficult to explain the reason for the delay in recovery of T3 and BIS. The context-sensitive half-life of magnesium has not, to our knowledge, been studied. Although the therapeutic and toxic serum magnesium concentrations are known, the clinical severity of the effects is not always correlated with the degree of hypermagnesaemia [8] and it is important to observe for early clinical signs of magnesium toxicity. In the anaesthetised patient, signs such as flushing, nausea, vomiting, somnolence, deep coma will clearly be missed. Cardiovascular monitoring in anaesthetised patients can only detect late signs of cardiac toxicity and the influence of other confounding factors is difficult to rule out. Finally, the authors have not elaborated on the cost effectiveness of the proposed regimen. We have made a calculation based on the data given in the study and the pricing given in the British National Formulary [3]. We estimated a total saving of only £0.80 over 5 h. The total cost of propofol, rocuronium and fentanyl used over the 5-h period for a patient with mean weight of 70 kg in their control group (group B) is £35.51. In the magnesium group (group A) with a mean weight of 65 kg the cost of the same drugs is £19.11, but there is an additional cost of £15.60 for the magnesium. We suggest that the use of this regimen will not be cost effective and would not be free from the risk of toxicity.