Ion-selective mini-electrodes have been widely employed to measure extracellular K+ and H+ during myocardial ischemia. However, the recent availability of this technology has not been accompanied by uniform fabrication, amplification, and calibration standards. In their fabrication, the chloride tips of Teflon-coated silver wires should be covered with a cellulose acetate-titanium dioxide sponge followed by a polyvinyl chloride (PVC)-valinomycin (K+) or PVC-tridodecylamine (H+) ion-selective membrane. Critical analysis of the nonworking electrodes using scanning electron micrographs has revealed membrane holes, membrane and sponge contamination, Teflon plaque, poor membrane-sponge-Teflon adhesion, and improperly applied or torn membrane. We have also found that signal amplification must have variable-gain filtration (0-1 Hz) with 0.5-pA input offset current and 10(12)-omega input resistance. Furthermore, in vitro calibration in 3 and 10 mM KCl (K+) or pH 8 and 6 buffer (H+) should produce a Nernstian slope +/- 5 or 10%, respectively, at 26 degrees C with a response time less than or equal to 50 ms, resistance greater than or equal to 10(12) omega, and drifts less than or equal to 1 mV/h. In vivo performance and calibration criteria (delineated for K+ only) include 1) transient response to bolus injections of KCl (0.12 mM/kg body wt) yielding peak amplitude changes of 2.5-3.0 mM, response times less than or equal to 10 s, and washout time constants less than or equal to 3 min, and 2) in vivo calibration to artificial independently confirmed systemic [K+] producing a Nernstian slope +/- 15% at 38 degrees C.(ABSTRACT TRUNCATED AT 250 WORDS)