The histopathologic effects of different doses of ethylcholine mustard aziridinium ion infused into the caudate-putamen complex or nucleus basalis were evaluated in rats. Although no non-specific tissue damage was observed at the lowest doses of ethylcholine mustard aziridinium ion examined--0.01 nmol in 1-microliter vehicle and 0.02 nmol in 2-, 5-, and 10-microliters vehicle in both the striatum and nucleus basalis--minimal but definite non-selective pathology, characterized by gliosis and loss of all neuronal elements in the region affected by the nitrogen mustard, was observed in both targets at a dose of 0.02 nmol 1 microliter and more severely at all doses containing 0.05 and 0.1 nmol ethylcholine mustard aziridinium ion. At doses of ethylcholine mustard aziridinium ion containing 0.2 nmol of the cytotoxin and greater amounts, non-specific cell loss in intact tissue and extensive cavitation became increasingly the most prominent histologic features of drug action. No statistically significant effects of ethylcholine mustard aziridinium ion on striatal choline acetyltransferase activities were found until doses of 0.4 nmol/1 microliter or greater were injected, concentrations of the cytotoxin at which appreciable non-specific pathology was also observed. Levels of dopamine in the caudate-putamen nucleus were reduced by comparatively greater amounts than choline acetyltransferase at doses of 2.5 nmol/2 microliters, 5.0 nmol/2 microliters and 10 nmol/2 microliters cytotoxin, but a significant effect of ethylcholine mustard aziridinium ion on striatal L-glutamate decarboxylase activity was found only at a dose of 10 nmol/2 microliters. As no dose of ethylcholine mustard aziridinium ion was found that reduced choline acetyltransferase without producing considerable non-specific tissue destruction, the usefulness of the cytotoxin in studying the behavioral and physiological consequences of selective cholinergic hypofunction in the brain must be questioned.