The nuclear run-on transcription assay is the only approach to measure the transcriptional activity of a given gene in its genuine structural and regulatory cellular context. However, serious problems in the interpretation of results can arise from the artificial activation of paused RNA polymerases during the transcription reaction, leading to false results with regard to the level and mode of gene regulation in vivo. We have used the example of the human proto-oncogene c-myc, which has previously been reported to be regulated by premature termination of transcription, to describe the problems and pitfalls in the interpretation of nuclear run-on experiments. We show here that activation of paused, elongation-incompetent polymerases in nuclear run-on experiments produces a strong transcription signal on c-myc exon 1 in cells which do not express c-myc steady-state RNA.