Timed assays in which GTP and GDP were separated and quantitated by HPLC were developed and used to study the metal activation of the mitochondrial and cytosolic isozymes of phosphoenolpyruvate carboxykinase purified from rabbit liver. These assays allowed both directions of catalysis to be studied under similar conditions and in the absence of coupling enzymes. The mitochondrial enzyme is rapidly inactivated by preincubation with Fe2+, as had been shown previously for the cytosolic isozyme. The greatest activation by Fe2+ was obtained by adding micromolar Fe2+ immediately after enzyme to form the complete assay mixture that also contained millimolar Mg2+. In the direction of synthesis of OAA from Pep, the K0.5 values for Mn2+ and Fe2+ were in the 3-7 microM range when a nonchelating buffer, Hepes, was used. The buffer used strongly affected activation by Fe2+ at pH 7.4; activation was eliminated in the case of phosphate and K0.5 increased several-fold over that obtained with Hepes when imidazole was used. In non-chelating buffer, the pH optimum was near 7.4 for both isozymes and for both directions of catalysis. However, the near optimal pH range extended below 7.4 for the direction of oxaloacetate synthesis while the range was above 7.4 for Pep synthesis. In the direction of oxaloacetate synthesis: (1) Both isozymes required the presence of micromolar Mn2+ or Fe2+ in addition to millimolar Mg2+ in order to shown significant activity. (2) Fe2+ was as effective an activator as Mn2+ at pH 7 and below. In the direction of Pep synthesis: (1) Micromolar Mn2+ was a much better activator than Fe2+ at the higher pH values needed for optimal activity in this direction. (2) With increasing pH, decreasing activation was obtained with Fe2+ while the activity supported by Mg2+ alone increased. The results demonstrate the potential for regulation of either isozyme of Pep carboxykinase by the availability of iron or manganese.