1. We have measured the ATP dependence of KATP channel activity, and the effect of various metabolites on this relationship, in inside-out membrane patches isolated from rat ventricular myocytes. 2. The inhibition of KATP channel activity by ATP could be described as a sigmoid function of [ATP] with a Hill coefficient (HATP) of 2 and a half-maximal inhibition at an ATP concentration (Ki, ATP) of 25 microM, in the presence of 0 mM, or 0.5 mM, total [Mg2+]. The non-hydrolysable ATP analogue, AMP-PNP, also inhibited the channel with Ki, AMP-PNP = 60 microM and HAMP-PNP = 2. 3. Acidosis caused a small, but significant, increase in Ki, ATP from 25 microM at pH 7.25 to 50 microM at pH 6.25, but phosphate and lactate were without effect (at 20 mM) on channel activity. 4. In the absence of ATP or Mg2+, ADP3- inhibited channel activity with Ki, ADP = 275 microM, and HADP = 1.2. Other purine and pyrimidine triphosphates, diphosphates and monophosphates also inhibited the channel with apparent order of inhibitory effectiveness ATP greater than AMP-PNP greater than ADP greater than CTP greater than GDP = AMP = ITP. 5. In the absence of Mg2+, but in the presence of 40 microM-ATP, channel inhibition by GTP, ITP, CTP, GDP, ADP or AMP was additive with inhibition by ATP. 6. In the presence of 0.5 mM-Mg2+ and 40 microM-ATP, inhibition by GTP, GMP and AMP was still additive with inhibition by ATP. The diphosphates ADP and GDP, however, paradoxically increased channel activity in the presence of ATP. This increase in channel activity appeared to result from a competitive increase in Ki, ATP, MgADP did not appear to cause any inhibition of channel activity. 7. We conclude that, in cardiac tissue, KATP channels are regulated by [ATP], and that this regulation is sensitive to other intracellular nucleotides, Mg2+, and pH, but not to phosphate or lactate. A simple, interactive two binding-site model is consistent with the nucleotide-dependent regulation that we observe.