Oximes are enzyme reactivators used in treating poisoning with organophosphorus cholinesterase (AChE) inhibitors. The oxime dose which can be safely administered is limited by the intrinsic toxicity of the substances such as their own AChE-inhibiting tendency. Clinical experience with the available oximes is disappointing. To meet this need, new AChE reactivators of potential clinical utility have been developed. The purpose of the study was to estimate in vitro both the intrinsic toxicity and the extent of possible protection conferred by established (pralidoxime, obidoxime, HI-6, methoxime, trimedoxime) and experimental (K-type) oximes, using diisopropyl-fluoro-phosphate (DFP) as an AChE inhibitor. The IC50 of DFP against human red blood cell AChE was determined ( approximately 120 nm). Measurements were then repeated in the presence of increasing oxime concentrations, leading to an apparent increase in DFP IC50. Calculated IC50 values were plotted against oxime concentrations to obtain an IC50 shift curve. The slope of this shift curve (tan alpha) was used to quantify the magnitude of the protective effect (nm IC50 increase per microm oxime). We show that, in the case of a linear relationship between oxime concentration and IC50, the binding constant K, determined using the Schild equation, equals IC50/DFP/tan alpha. Based on the values of tan alpha and of the binding constant K, some of the new K-oxime reactivators are far superior to pralidoxime (tan alpha = 0.8), obidoxime (1.5), HI-6 (0.8), trimedoxime (2.9) and methoxime (5.9), with K-107 (17), K-108 (20), and K-113 (16) being the outstanding compounds.