Quantitative measurement of BK virus DNA (Q-BKDNA) has been used for the early diagnosis and monitoring of BK virus-associated nephropathy (BKVAN). This study was designed to determine the BKDNA cutoff for the diagnosis of BKVAN. Between June 2005 and February 2007, 64 renal transplant recipients taken renal biopsies due to renal impairment submitted plasma and urine for Q-BKDNA. Eight BKVAN patients (12.5%) had median viral loads of 6.0 log10 copies/mL in plasma and 7.3 log10 copies/mL in urine. Among 56 non-BKVAN patients, 45 were negative for Q-BKDNA; 4 were positive in plasma with a median viral load of 4.8 log10 copies/ mL, and 10 were positive in urine with a median viral load of 4.8 log10 copies/mL. Receiver operating characteristic curve analysis showed that a cutoff of 4.5 log10 copies/mL in plasma and a cutoff of 5.9 log10 copies/mL in urine had a sensitivity of 100% and a specificity of 96.4%, respectively. A combined cutoffs of 4 log10 copies/ mL in plasma and 6 log10 copies/mL in urine had better performance with a sensitivity of 100% and a specificity of 98.2% than each cutoff of urine or plasma. QBKDNA with the combined cutoffs could reliably diagnose BKVAN in renal transplant recipients.