We have investigated prospectively the efficacy of two nonpharmacologic relaxation techniques in the therapy of anxiety. A simple, meditational relaxation technique (MT) that elicits the changes of decreased sympathetic nervous system activity was compared to a self-hypnosis technique (HT) in which relaxation, with or without altered perceptions, was suggested. 32 patients with anxiety neurosis were divided into 2 groups on the basis of their responsivity to hypnosis: moderate-high and low responsivity. The MT or HT was then randomly assigned separately to each member of the two responsivity groups. Thus, 4 treatment groups were studied: moderate-high responsivity MT; low responsivity MT; moderate-high responsivity HT; and low responsivity HT. The low responsivity HT group, by definition largely incapable of achieving the altered perceptions essential to hypnosis, was designed as the control group. Patients were instructed to practice the assigned technique daily for 8 weeks. Change in anxiety was determined by three types of evaluation: psychiatric assessment; physiologic testing; and self-assessment. There was essentially no difference between the two techniques in therapeutic efficacy according to these evaluations. Psychiatric assessment revealed overall improvement in 34% of the patients and the self-rating assessment indicated improvement in 63% of the population. Patients who had moderate-high hypnotic responsivity, independent of the technique used, significantly improved on psychiatric assessment (p = 0.05) and decreased average systolic blood pressure from 126.1 to 122.5 mm Hg over the 8-week period (p = 0.048). The responsivity scores at the higher end of the hypnotic responsivity spectrum were proportionately correlated to greater decreases in systolic blood pressure (p = 0.075) and to improvement by psychiatric assessment (p = 0.003). There was, however, no consistent relation between hypnotic responsivity and the other assessments made, such as diastolic blood pressure, oxygen consumption, heart rate and the self-rating questionnaires. The meditational and self-hypnosis techniques employed in this investigation are simple to use and effective in the therapy of anxiety.