A method of therapeutic monitoring of cyclosporine by using area under the curve (AUC) has been previously proposed. However, different mathematical methods of calculating AUC may produce different results. We compared three methods of calculating whole blood 12-h AUC of cyclosporine A (CsA) at steady state in 16 pediatric renal allograft recipients. The linear trapezoidal method tended to significantly overestimate AUC when compared with a method combining linear and log trapezoidal methods, or the Lagrange technique combined with the logarithmic trapezoidal method, producing mean differences of 13.8 ng*h/ml [95% confidence interval (CI), 7.3-20.3] and 12.8 ng*h/ml [95% CI, 7.3-18.3], respectively. However, these differences appear to be of little clinical significance because they comprised < 1% of the calculated AUC value. The calculated AUCs by the three methods produced values with similar means and overlapping 95% CI. These data suggest that use of any of these three mathematical methods, when used to calculate CsA AUC for the determination of average steady-state concentrations and dose rate calculations, will produce similar results.