This article reviewed and analyzed 39 studies on self efficacy theory applied to health related behavior. The following analysis was done: 1) study subjects 2) measurement tools 3) analysis according to the type of research design(intervention research, explanatory research). Some findings are summarized as follows: The study subjects were both healthy people in various developmental stages and patients with various illnesses. The health related behaviors examined in the studies were also various including exercise, smoking cessation, self care behaviors, etc. The measurement of self efficacy was done with specific tools in most studies. In the tools, activities that measured the health behavior domain were listed according to increasing difficulty or contexual arrangement or in combination of both of them. The analysis of 17 intervention research studies showed that generally the intervention program increased the self efficacy level of subjects and then the increased strongly self efficacy influenced behavioral changes. Most studies used more than one intervention method for increasing the self efficacy level. These were derived from sources of self efficacy suggested by Bandura. The analysis of 21 explanatory research studies showed that self efficacy strongly influenced behavior change and persistence. The major independent variable to affect the self efficacy was performance accomplishment in the past. Self efficacy explained more of the variance in health related behavior when it was applied with the variables in the health belief model, health promotion model, and reasoned action theory. On the basis of the above findings, the following suggestions are made: 1. For a desirable research design, self efficacy should be the intervening variable. That is, desirable designs would include intervention-self efficacy-behavior in intervention research studies and antecedent-self efficacy-behavior in explanatory research studies. 2. More prospective, longitudinal studies are needed to test the effect of self efficacy on persistence in health related behavior. 3. Studies comparing the effects of intervention methods are needed for each health related behavior, subject group, and context. 4. It is necessary to develop a reliable, valid measurement tool for self efficacy for each health related behavior. 5. Studies to differenciate the effect of self efficacy from that of outcome expectation on the health related behavior are necessary. 6. The antecedents of self efficacy should he investigated further.