BACKGROUND: The objective of this study was to use information from a donor to establish the lowest possible serum creatinine (SCr) of the recipient as a means of identifying early graft dysfunction. METHODS: We analyzed retrospectively 58 pairs of living donors and recipients. The lowest possible SCr was calculated from four different formulae derived from Cockcroft-Gault formula: Ax(140-recipient age)xrecipient weight/donor GFR (A, women: 1.04, men: 1.23). Donor GFR was represented by the following values (a) mSCrD0: measured pretransplant GFR for both kidneys, (b) mSCrSKD0: single transplanted kidney GFR on day 0, (c) eSCrD0: estimated GFR based on SCr at day 0, or (d) eSCrD30: on day 30. These resulting estimated SCr were tested for correlation coefficient, bias, precision, and accuracy in predicting the lowest observed recipient SCr during the first year posttransplant. RESULTS: The lowest possible SCr was 80+/-22 micromol/L for mSCrD0, 79.1+/-23 micromol/L for mSCrSKD0, 83+/-27 micromol/L for eSCrD0, and 115+/-22 micromol/L for eSCrD30. Mean values for lowest possible SCr correlated significantly with the lowest observed SCr for all four formulae. The eSCrD0 formula showed the best correlation (r=0.47), the smallest positive bias (20.5 micromol/L), the highest precision (21.9 micromol/L), and the second highest percentage of predicted values that fell within 30% of the observed SCr (69%). CONCLUSION: The use of the Cockcroft-Gault derived formula with eSCrD0 may be a useful tool to detect early discrepancy between observed and lowest possible SCr value. This could help to identify patients who may require invasive investigations.