The purpose of this dissertation was to test the significance of objective l y measured imagery ability on the learning of self-controled relaxation of autonomic nervous system activity. Imagery is discussed in terms of its interaction with Autogenic vs. Jacobsonian methods of training clinical relaxation. Thirty-six female subjects from a college population, representing extreme highs and lows on "spatial ability" tests were given a series of three six-session sequences of Baseline, Treatment 1, and Treatment 2, which contained silent relaxation as a control, plus Jacobsonian and Autogenic relaxation. High and low spatial ability subjects were divided into split groups (A & B) which were given Jacobsonian and Autogenic relaxation treatment in different sequence orders. Skin temperature biofeedback was used to monitor the little fingers on both hands as a general indicator of autonomic clinical relaxation. Mean temperature; temperature change within sessions; and temperature change between sessions, were analyzed by different treatment periods and spatial ability groups. The data from these groups were analyzed using an ANOVA design. There were no significant differences in mean temperature data. A nearly significant two-way interaction was found between imagery ability and treatment order during Autogenic training. Also a significant interaction was found in skin temperature change between sessions for, "Sensory" vs. "Intuitive" personality types, and a nearly significant difference for Autogenic vs. Jacobsonian treatment. It was concluded that Jacobsonian training was generally more effective than Autogenic training for inducing vascular relaxation in both high and low imagery subjects. Also it was found that Sensory perceptual types are significantly more stable in terms of day to day skin temperature variation during relaxation training, than are Intuitive perceptual types.