We evaluated total and ionized plasma magnesium levels, and erythrocyte and platelet magnesium concentrations from two groups of renal transplant recipients treated either with cyclosporine, azathioprine and prednisolone (group CAP, n = 8) or with azathioprine and prednisolone (group AP, n = 13), and in a group of age- and sex-matched healthy subjects (n = 10). Reduced plasma (total and ionized), erythrocyte and platelet magnesium concentrations were found in both CAP and AP groups with respect to controls (CAP: total plasma Mg median 0.61 vs 0.86 mmol/L, p < 0.01, ionized plasma Mg median 0.43 vs 0.58 mmol/L, p < 0.001, erythrocyte Mg median 2.18 vs 2.56 mmol/L, p < 0.05, platelet Mg median 1.75 vs 2.84 mmol/10(8) cells, p < 0.001; AP: total plasma Mg median 0.62 vs 0.86 mmol/L, p < 0.01, ionized plasma Mg median 0.48 vs 0.58 mmol/L, p < 0.001, erythrocyte Mg median 2.30 vs 2.56 mmol/L, p < 0.05, platelet Mg median 1.75 vs 2.84 mumol/10(8) cells, p < 0.001), while no difference was found between the two groups of transplant recipients as regards plasma and intracellular magnesium levels. Magnesium fractional excretion was higher in transplant recipients than in the control group (Mg fractional excretion median AP 18.6 per cent and CAP 12.8 per cent vs controls 3.5 per cent), whereas no difference was found between patients and control subjects for urinary magnesium 24h excretion. Moreover, in the whole group of transplant recipients (n = 21), urinary magnesium showed an inverse correlation with platelet (rs = -0.54, p < 0.05) and ionized plasma magnesium (rs = -0.48, p < 0.05), and time after transplantation showed a negative correlation with platelet magnesium concentrations (rs = -0.73, p < 0.001), and a direct correlation with fractional magnesium excretion (rs = 0.53, p < 0.05). Finally, a direct relationship between platelet magnesium and ionized plasma magnesium was also detected in the whole group of transplant recipients (rs = 0.47, p < 0.05). Both intraplatelet magnesium depletion and ionized plasma magnesium reduction induced by immunosuppressive therapy could be involved in the increased risk from atherosclerotic disease in renal transplant recipients.