A considerable proportion of patients with chronic hepatitis C who achieve a virologic end-of-treatment response relapse after discontinuation of therapy. It is conceivable that polymerase chain reaction (PCR)-based assays with a lower detection limit of 100 to 1, 000 hepatitic C virus (HCV) RNA copies/mL are still too insensitive to detect residual viremia. End-of-treatment serum samples of 47 patients with a virologic relapse according to results of qualitative PCR assays (Amplicor HCV; Roche Molecular Systems, Mannheim, Germany) were tested by transcription-mediated amplification (TMA), an isothermal, autocatalytic target amplification method that has the potential to detect less than 50 HCV RNA copies/mL. Virologic sustained responders (n = 59) and nonresponders (n = 49) served as controls. In end-of-treatment serum samples of virologic sustained responders and nonresponders an almost complete concordance between PCR and TMA results was observed (98%). However, HCV RNA was detectable by TMA in end-of-treatment serum samples from 16 of 25 relapse patients (64%) who were HCV-RNA-negative according to Amplicor HCV version 1.0 (lower detection limit 1,000 copies/mL) and in 8 of 22 patients (36%) who were HCV-RNA-negative according to Amplicor HCV version 2.0 (lower detection limit 100 copies/mL). End-of-treatment alanine transaminase (ALT) levels of sustained virologic responders and TMA-negative relapsers were similar, whereas a trend toward higher ALT values was observed in TMA-positive relapsers compared with sustained virologic responders (P = 0.09). In conclusion, HCV RNA can be detected at the end of treatment by TMA in a considerable proportion of patients who were classified as virologic end-of-treatment responders with a subsequent virologic relapse according to PCR-based methods.