Direct biofeedback of blood pressure was compared with frontal EMG biofeedback and with self‐instructed relaxation for the treatment of essential hypertension in a controlled group outcome study. Patients were followed up for four months after the end of treatment. Generalization of treatment effects was assessed through pre‐ and posttreatment measurements of blood pressure under clinical conditions in a physician's office. There were no significant reductions in diastolic blood pressure. The systolic blood pressure (SBP) of the patients receiving blood pressure biofeedback decreased 8.1 mm mercury (p= 0.07) and the SBP of the patients in the relaxation condition decreased 9.5 mm mercury (p= 0.05). In the generalization measures, there were significant reductions in SBP for the relaxation group. The results are discussed in terms of the general lack of replicability within the area of biofeedback treatment of hypertens