Abstract Pyruvate-supported 11β-hydroxylation of 11-deoxycorticosterone (DOC) via cytochrome P-450 reductase chain in incubated rat adrenal mitochondria was maximal when either traces of oxaloacetate (OAA) or 2 mM ATP were added to the incubation medium. This showed that reducing equivalents formed at the pyruvate dehydrogenase level as well as those derived from Krebs-cycle activity were needed for maximal corticosterone (B) formation from DOC. These findings were confirmed in experiments with whole cells isolated from rat adrenals designed to show the possible pathways of intramitochondrial NADPH generation for steroid hydroxylations. Whereas citrate, isocitrate, succinate and malate were not metabolized because of their impermeability to the plasma membranes of the cells, both [1- 14C] and [2- 14C]-pyruvate were efficiently utilized by the mitochondria of the cells for B formation from DOC. Whereas arsenite completely inhibited the pyruvate- 14C supported B formation, comparison of the data obtained with two inhibitors, 2,4-dinitrophenol (2,4-DNP) and fluorocitrate, showed that for the same per cent inhibition of 14CO 2 production the inhibition of 11βJ-hydroxylation of DOC was greater with 2,4-DNP than with fluorocitrate. It is concluded that operation of the Krebs-cycle is essential for optimizing the production of reducing equivalents needed for 11β-hydroxylation to occur. It is also concluded that the rate of oxidation of pyruvate in the mitochondria of the cells and the activity of the Krebs-cycle is dependent on availability of OAA. When ATP levels were reduced by the uncoupling effect of 2,4-DNP resulting in a decreased production of OAA via ATP-requiring pyruvate carboxylase, a concomitant inhibition in production of reducing equivalent occurred which led to a lack of B production from DOC. Mitochondria isolated from the steroid-producing Snell adrenocortical carcinoma 494 cells (P 2T) were unable to oxidize pyruvate and several Krebs-cycle substrates. Succinate and α-glycerol phosphate which both were oxidized via their respective P 2T flavoprotein linked enzymes, supported high rates of O 2 uptake but had little effect on DOC conversion into B. Electron micrographs of P 2T when compared to P 2C from normal rat adrenals showed that their ultrastructure was markedly altered. The number of mitochondria per cell in the tumor tissue was also considerably lower than that found in normal adrenal cells. Many of the P 2T were elongated, had lamillar-shaped cristae and showed a much less uniformed shape than P 2C which are usually oval and contain packed vesicular cristae. Adrenodoxin reductase activity, cytochrome P-450 and NADP nucleotide levels in P 2T were also lower than those in P 2C which partly accounted for the low conversion of DOC into corticosterone. Because (1) mitochondrial flavoprotein-linked a-glycerol phosphate dehydrogenase activity in P 2T was 10 times higher than that found in P 2C, (2) an active NAD-linked α-glycerol phosphate dehydrogenase in the cytosol of the carcinoma cell was found and (3) pyruvate is not utilized by P 2T or whole cancer cells, this suggested the possibility of an α-glycerol phosphate shuttle operating in the tumor cells. The purpose of the shuttle might be to provide some of the cellular ATP required in this rapidly growing cancer tissue.