Transcription in heterochromatin seems to be an oxymoron--surely the 'silenced' form of chromatin should not be transcribed. But there have been frequent reports of low-level transcription in heterochromatic regions, and several hundred genes are found in these regions in Drosophila. Most strikingly, recent investigations implicate RNA interference mechanisms in targeting and maintaining heterochromatin, and these mechanisms are inherently dependent on transcription. Silencing of chromatin might involve trans-acting sources of the crucial small RNAs that carry out RNA interference, but in some cases, transcription of the region to be silenced seems to be required--an apparent contradiction.