In vitro, for animal cells generally, butyrate at millimolar concentrations is an inhibitor of growth. In vivo, however, colonocytes are able to grow in the environment of about 20 mM butyrate produced by bacterial fermentation on the luminal side of the colonic epithelium. Anin vivoincrease of the butyrate supply results in growth stimulation of cells in the colonic crypts. This discrepancy, namely, that in cell cultures butyrate is an inhibitor of growth, whereasin vivoit has a trophic effect, is the so called in vivo paradox of butyrate. In the present review it is pointed out that butyrate is an inhibitor of histone deacetylases and there is sufficient evidence for hyperacetylation being the mechanism of thein vitrogrowth-inhibiting effect of butyrate. As within animal cells hyperacetylation has to occur at a certain butyrate concentration (1–10 mM), it is postulated that the in vivo lack of inhibition and 'paradoxical' stimulation of growth is a result of a low intracellular steady state concentration of butyrate in the lower layers of the crypt in spite of the much higher butyrate concentration on the luminal side. As butyrate is the preferential source of energy for colonocytes, thein vivotrophic effect is not paradoxical, when in spite of an increase of the butyrate concentration in stool, the intracellular butyrate concentration of intestinal epithelial cells still remains below the inhibiting level. For mature non-dividing colonocytes which are programmed for apoptosis, there is no difference between the observations madein vitroor in vivo. Furthermore, recent developments are discussed which suggest that cyclo-oxygenase-2 may play an essential role in colonic carcinogenesis. Cyclo-oxygenase- 2 is found to be expressed in most colorectal carcinomas, but not in normal non-transformed intestinal epithelial cells (DeWitt and Smith, 1995). Cyclo-oxygenase-2 overexpression makes intestinal epithelial cells resistant to butyrate-induced apoptosis (Tsujii and DuBois, 1995). This escape from butyrate-induced apoptosis appears to be an essential prerequisite for the development of colorectal cancer and suggests a functional role of butyrate in growth, differentiation and programmed cell death of colonic epithelial cells.