SummaryA variety of reports describe shifts in the environment which cause a corresponding change in the measured linking number of piasmid DNA isolated from bacterial celis. This change in iinking number is often attributed to a change in superhelicai density. This, coupled with the observation that transcription is often dependent upon the superhelicai density of the DNA template seen^n vitro, has led to the suggestion that superhelicai density may control expression of certain genes. However, since many environmental changes could, in principle, influence DNA twist itself, then the measured differences in linking number, hLk, may simply be a consequence of variation in twist according to the relationship bLk = hTw + hWr, where bTw and bWr are changes in twist and writhe, respectively. In fact, we show that when an environmental change causes a change in the helical pitch of the DNA, and if the superhelicai density of DNA is regulated to remain constant according to the homeostatic model of Menzel and Gellert, then bLk = bTw. We have found that there are a number of published reports describing variation in promoter activity as a function of linking number that can be explained by considering twist. We suggest that there are classes of a^° promoters whose ability to be recognized by RNA polymerase is exquisitely sensitive to the relative orientation of the -35 and -10 regions, and environmental conditions can control this relative orientation by changing DNA twist. The recA and proU promoters which are activated by cold shock and osmotic shock, respectively, behave as if they are twist-sensitive promoters. Consideration of DNA twist can also account for the change in activity of a number of other promoters when they are placed