In clinical practice p-creatinine is used to estimate changes of GFR. Generally, it is believed that recovery of p-creatinine within 10% of initial baseline allows exclusion of relevant nephrotoxic changes. We evaluated whether recovery of GFR after discontinuation of CyA therapy can be adequately predicted by measuring p-creatinine alone. Fifty-four allogenic BMT patients were followed up by p-creatinine and classical inulin clearance (GFR) before BMT and 1, 3, 6, 12, 18, 24 months after BMT. A total of 10 patients fulfilled following three criteria: (1) 24 months total follow-up time; (2) at least 12 months follow-up after discontinuation of CyA therapy (3) no trimethoprim or cimetidine comedication at time of clearance measurement. Time after CyA withdrawal varied between 13 and 21 months (mean +/- standard deviation, 17 +/- 2 months); mean duration of CyA therapy was 8 +/- 2 months (minimum: 3 months, maximum: 11 months). After at least 12 months of CyA stop mean p-creatinine returned to baseline values. In contrast, mean GFR remained about 20% below baseline (paired sample Wilcoxon-test P < .02). Neither creatinine excretion nor body weight nor creatinine clearance changed significantly between baseline and 24 months after BMT. Follow-up of p-creatinine after CyA stop can overestimate the recovery of GFR. A 20% loss of GFR may remain unrecognized. We speculate that this phenomenon is due to tubular hypertrophy in the recovery phase.