The objective of this study was to compare the sensitivity of three different reverse transcriptase-polymerase chain reaction (RT-PCR) based tests, for detection of GB virus C (GBV-C) RNA. One commercial and two 'in house' RT-PCR assays were employed in the testing of serum samples from 114 chronic hepatitis C infected individuals. A part of the 5' non-coding region (5'NCR) of the GBV-C genome was amplified by the GBV-C LCx assay (Abbott) and one of the 'in house' RT-PCR tests. In the other 'in house' RT-PCR a segment of the NS5B region was amplified. The 'in house' assays included the use of internal controls that were co-amplified with use of the same outer PCR primers as the virus targets. The GBV-C LCx from Abbott and 5'NCR 'in house' PCR tests detected 28 and 27 GBV-C positive individuals, respectively. The sample positive only in the LCx test was confirmed by the 'in house' 5'NCR RT PCR using an increased virus input. In comparison, the NS5B 'in house' PCR test detected 24 of the GBV-C positive samples. One sample showed no amplification of internal controls/virus target in the 5'NCR 'in house' PCR and another samples was amplification negative in the NS5B PCR. The PCR assays with primers from the 5'NCR of the virus genome e.g. the GBV-C LCx, were more sensitive compared with RT-PCR using primers from the NS5B region. The GBV-C LCx seemed to be the most sensitive and robust assay. Internal controls included in the 'in house' assays identified two samples with failure of the amplification.