Studies of behavioral traits of the "Arkansas Line of nervous pointer dogs" are well documented (Murphree et al. 1969). Demonstrations of drug effects on these traits, as measured on the standardized behavior test and expressed as "Morbidity Score," have been disappointing, despite other evidence of effectiveness of drug treatments such as modification of operant performance (Murphree et al. 1974c). A simple reproducible test with high interrater reliability is presented here. Subjects were nine nervous dogs that received placebo, pimozide, and chlordiazepoxide in a double-blind crossover design. Both drugs significantly attenuated "human avoidance," a nervous trait. Placebo had no effect. The chlordiazepoxide effect was maximal on the first day of administration, returning to baseline three to four days after the last dose. The pimozide maximal effect was on day 4, with return to baseline six to nine days after the last dose. In addition to the aberrant behavior of the nervous pointer, hyperresponsiveness to dopaminergic stimulation and increased levels of L-DOPA in the cisternal spinal fluid were observed.