Most patients having leukemia suffer severely from emotional turmoil due to the generalized perception that cancer will be fatal. The conventional chemotherapy results in side effects such as severe bone marrow depression which interfere with self-care management, vital for improvement in their condition. Bandura's theory of the self-efficacy suggests that self-efficacy can be enforced by performance attainments, vicarious experiences, verbal persuasion and the release of emotional arousal. Self-efficacy can be enforced by a program of vicarious experiences and verbal persuasion, while the emotional arousal can be relieved through a hope promotion program. If once self-efficacy increases, the patient's self-care behaviors and the quality of life will also increase. The purpose of this study was to empirically test the effects of a program, to promote self-efficacy and hope, on self-care behaviors and quality of life in patients having leukemia. In this study, three types of approaches to enhance self-efficacy and hope were used : 1) a 20-minute long slide/tape for vagarious experiences ; 2) a 10-minute long telephone call coaching for verbal persuasion ; and 3)two booklets for information about the symptoms of leukemia and treatment modalities and hope promotion. Thirsty one patients were recruited in the experimental group and 29 in the control group with a nonequivalent pretest-posttest design. The subjects were patients with leukemia undergoing chemotherapy. Sherer and Maddux's self-efficacy scale, Nowotny's hope scale, and Padilla's quality of life scale were employed with some modifications. A self-care behavior scale was developed by the researchers. Statistical analyses including paired t-test, Chi-square, ANCOVA and ANOVA, were used. The results are as follows : The levels of self-efficacy, self-care behavior and quality of life were higher in the experimental group than in the control group after four weeks of intervention(F=28.71, P=.0001 ; F=63.35, P=.0001 ; F=16.57, P=.0001). After ten weeks of intervention, all of the dependent variables(self-efficacy, self-care behavior, hope and quality of life) in the experimental group were higher than in the control group(F=74.12, P=.0001 ; F=108.34, P=.0001 ; F=13.11, P=.001 ; F=43.52, P=.0001). In conclusion, self-care behavior and quality of life increased mainly through an increase an increase in self-efficacy, while increases in hope took more time and effort.