Transcription of vesicular stomatitis virus (VSV) in Vero cells was confined to the synthesis of parentally-derived mRNA (primary transcription) by the use of cycloheximide and/or a ts mutant, G41(IV), at a non-permissive temperature (40 degrees C). More transcripts accumulated in the presence of cycloheximide than in its absence. This so-called "cycloheximide effect" results from higher rates of virus transcription sustained for longer periods of time. The rate of VSV transcription initially increases linearly for 1 to 2 h after injection. Interferon reduces this rate (congruent to fourfold with 50 units/ml interferon) irrespective of the presence or absence of cycloheximide. The VSV mRNA transcripts synthesized in mock- or interferon-treated cells were equal in size and had an equivalent half-life of 17 h at 40 degrees C. It seems likely that once transcription is initiated in interferon-treated cells, it is completed successfully. Since interferon reduces the rate of early VSV primary transcript synthesis to below that achieved in the presence of cycloheximide, we conclude that interferon has an effect on transcription beyond that attributable solely to protein synthesis inhibition. We postulate that interferon decreases the probabiltiy of initiating virus transcription. Virus mRNA escaping this facet of interferon action may then encounter other facets such as post-transcriptional modification and/or inhibition of translation. However, the mandatory sequence of primary transcription leads to primary translation for negative-strand viruses like VSV dictates that the overall inhibitory effect of interferon on translation would derive in part from this prior inhibition of transcription. Thus, to apply the term "primary effect" to one particular facet of interferon action may not always be meaningful.