A recently developed reverse transcription-nested polymerase chain reaction (RT-nPCR) method for rubella virus (RV) RNA detection was assessed in a series of African green monkey kidney (AGMK) cell cultures inoculated with clinical samples from patients with suspected RV infection. Results were compared with those of conventional virus isolation/identification. The assay included an internal control of amplification consisting of a synthetic RNA molecule mimicking the RV E1 target sequence. A semiquantitation of RV RNA was achieved by comparing the relative band intensity of internal control and RV E1 fragment. RT-nPCR was positive in 15/16 (94%) RV isolation-positive cultures and in 12/60 (20%) RV isolation-negative cultures. All 27 cell cultures positive by RT-nPCR had been inoculated with clinical samples taken from patients with ascertained RV infection or given RV vaccination and consisted of cells harvested 1-10 days after primary inoculation of clinical samples. No RV RNA was found in any of the cell cultures inoculated with 14 clinical samples from 6 patients in whom RV infection was excluded. When considering the time-course of RV infection, it was found that RV RNA could be detected as early as 4 days p.i. in 10/21 (48%), and by 7-10 days p.i. in 27/28 (96%) cell cultures, whereas by the same time RV was isolated in only 7/16 (44%) cell cultures. Semiquantitation showed that: i) viral RNA amount progressively increased with time; ii) cell cultures containing very low levels of viral RNA one week p.i. either required a few blind passages for virus recovery or remained negative for RV isolation. Finally, PCR inhibitors were found in 10/164 (6%) cell cultures examined. In conclusion, RT-nPCR proved to be very sensitive and very specific and greatly reduced RV detection time in inoculated cell cultures.