As DNA gyrase is the only enzyme to supercoil DNA actively, we address here the question of whether it does play the expected dominant role in controlling the level of DNA supercoiling and growth rate in Escherichia coli. We modulated the expression of DNA gyrase around its wild-type level, and measured the effect on plasmid supercoiling and growth rate. As both the activity and the transcription rate of DNA gyrase are sensitive to DNA supercoiling we distinguish two types of control (with control defined as the percentage change observed on a 1% modulation of a parameter). The first type of control, here named inherent control, quantifies the effect of a sustained modulation of the transcription rate of gyrase. At its wild-type expression level this inherent control exerted by DNA gyrase on growth rate was very low, i.e. c(μ)(gyrase) = 0.05 - 0.00, as was the inherent control on DNA supercoiling, c(aLk)(gyrase) = 0.2. The second type of control, here named global control, quantifies the effect of a change in gyrase activity whilst allowing the cell to respond by readjusting gyrase transcription. Both types of control are linked via the sensitivity of gyrase transcription to DNA supercoiling, as determined from the inherent control by gyrase of the gyrase promoter activity using a chromosomal gyrB::lacZ fusion. As expected, the latter control was negative, but small, i.e. c(gyr promoter)(gyrase)= -0.3. The global control by gyrase of active linking number was 0.1. These results show that although gyrase is an essential enzyme it does not have a high control, on either growth rate or DNA supercoiling. Homeostatic regulation of physiological DNA structure appears to dominate. At low degrees of DNA supercoiling, the control by DNA gyrase and by the other topoisomerases is much stronger.