A prospective study of 128 adult cardiac surgical patients was undertaken in order to quantify net magnesium loss and its relationship to serum magnesium levels and postoperative problems, particularly arrhythmias. Peri-operative magnesium flux on the first day was calculated from the administered magnesium (in cardioplegia solution and intravenous infusion) and urinary magnesium loss. Magnesium input ranged from 24 mmol to 40 mmol, resulting in a net magnesium gain in 94% of patients. Hypomagnesaemia, identified in 34% of patients pre-operatively and 30% of patients postoperatively, had no significant correlation with the measured peri-operative magnesium flux or the electrocardiograph corrected-QT interval. Fifty-three patients developed postoperative arrhythmias, but there was no significant correlation with the serum total magnesium concentration, or with the peri-operative change in serum magnesium level, magnesium flux, or QT interval. The data suggest that serum total magnesium is not a useful measurement upon which to base preventive or therapeutic measures in cardiac surgical patients.