The site of S-phase DNA synthesis has been the subject of recurring controversy. All recent evidence supporting a site fixed to some nuclear sub-structure is derived from studies in which cells or nuclei have been extracted in hypertonic salt concentrations. The controversy centres on whether the resulting nuclear matrices or cages have counterparts in vivo or are simply artefacts. Using isotonic conditions throughout the isolation and analytic procedures we have now reinvestigated the site of replication. Cells are encapsulated in agarose microbeads and lysed to leave encapsulated nuclei which are nevertheless completely accessible to enzymes. After incubation with endonucleases, most chromatin can be electroeluted from beads: however, nascent DNA and active DNA polymerase remain entrapped. Since chromatin particles containing DNA the size of 125 kbp can electroelute, we conclude that the polymerizing complex is attached to a nucleoskeleton which is too large to escape. We have also studied various artefacts induced by departure from isotonic conditions. Perhaps surprisingly, the hypotonic conditions used during isolation of nuclei by conventional procedures are a significant source of artefact.