The dose-response curve of procaine or lignocaine (lidocaine) added to the St. Thomas' Hospital cardioplegic solution was investigated with an isolated working rat heart preparation. In the absence of any cold cardioplegic protection, hearts failed to recover after as little as 30 minutes of ischemia. A single infusion (20 degrees C) of the basic St. Thomas' Hospital cardioplegic solution allowed hearts to recover to 60% or more of their preischemic control aortic flow after a 120 minute period of ischemia. Addition of procaine to the cardioplegic solution either increased or reduced the apparent protective properties of the solution with a bell-shaped dose-response curve being obtained. The optimum procaine concentration was 0.05 mM/L. At this concentration the protection afforded by the St. Thomas' Hospital solution was increased by up to two thirds. Substitution of lignocaine for procaine resulted in a similar dose-response curve with its optimum also at 0.05 mM/L. If a similar optimum exists for the human heart, the doses in current clinical use would appear to be too high. These results argue for determining the dose-response characteristics of all substances used in cardioplegic solutions.