Summary: Isolated Langendorff-perfused rat hearts set up to allow measurement of mechanical myocardial function (according to Bardenheuer and Schrader) were subjected to 60 min of global ischemia (gI). The hearts were perfused with a modified Krebs-Henseleit solution (KHS; control group) or KHS with desipramine (DMI group) 100 nM. Extracellular concentrations of K+, Na+, and H+ were determined using ion-selective membrane polyvinyl chloride (PVC) minielectrodes. The noradrenaline (NA) release was estimated by analyzing the NA overflow during a 10-min reperfusion period, using high pressure liquid chromatography (HPLC). Control hearts and DMI-treated hearts showed no differences in mechanical performance prior to gI. In control hearts, the onset of gI resulted in a rapid triphasic [K+]e accumulation (an increase from 4.25 ± 0.25 to 8.76 ± 0.37 mmol/ L during the first 3.5 min, a decrease to 8.06 ± 0.34 mmol/ L until the seventh min, and a further increase to 33.0 ± 2.42 mmol/L at 60 min), a rapid increase in [H+]e (45.9 ± 2.3 to 1314.5 ± 214 nmol/L during 23 min), and a slight initial increase of [Na+]e (138.8 ± 0.4 to 141.9 ± 0.8 mmol/L within 8 min) followed by a long-lasting decrease (119.1 ±4.5 mmol/L at 60 min). Treatment with DMI had only slight effects on ionic redistribution during the course of ischemia. The [K+]e and [H + ]e accumulation tended to be diminished, whereas [Na+]e values were slightly higher in the course of gI compared to the control group. The ischemia-induced NA overflow was found to be significantly lower in DMI-treated hearts (148.9 ± 19.1 versus 226.0 ± 25.0 ng/g heart/5 min in control hearts), but in both experimental groups the estimated values of extracellular NA after 60 min of gI were in a range known to cause cellular necrosis (> 1 μmol/L). In conclusion, the results of this study indicate that (a) the isolated rat heart preparation represents a suitable model to determine changes of extracellular ion concentrations during ischemia, (b) the slight influences of DMI on the ion redistribution during gI are mainly due to an unspecific membrane stabilizing effect, and (c) therefore cannot explain the reduction of NA release following DMI application