The aim of this study was to compare the sensitivity and specificity of conventional procedures (in-house one-stage polymerase chain reaction (PCR) and in-house nested PCR) and of new technologies (rTth DNA polymerase (Amplicor), branched-DNA, NASBA (nucleic acid amplification system)) for the qualitative detection of hepatitis C virus (HCV) RNA in serum of HCV-infected individuals. Serum samples from 37 anti-HCV-positive individuals (15 with a normal alanine aminotransferase (ALT) level, 22 with an elevated ALT level) and 10 anti-HCV-negative individuals as negative controls were studied. A second panel, including 9 diluted serum samples (from 1/10 to 1/100,000) was constituted to establish the differences of sensitivity of the 5 procedures with small quantities of HCV RNA in the serum. The anti-HCV-positive individuals with elevated ALT gave positive results with all 5 procedures. In patients with a normal ALT level, the assays with the highest sensitivity were Amplicor, NASBA and nested RT-PCR, followed by one-stage RT-PCR, then branched-DNA. One false-positive result was observed with Amplicor, and two with in-house nested PCR. On diluted samples, Amplicor, NASBA and nested PCR appeared more sensitive than one-stage PCR and branched-DNA. It is concluded that new procedures have satisfactory sensitivity and specificity and could advantageously replace the conventional PCR procedures for the routine qualitative detection of serum HCV RNA.