The purpose of this study was to explain the relationship between smoking status, smoking behavior and hardiness in university students in Gangnung City. The subjects were a convenience sample of 315 students. The data were collected by a questionnaire given to the students between May 22 to June 2, 2000. An instrument developed by Akers & Gang(1996) and translated by Sohn, Jung-Nam(1999) was used in this study to measure definition of smoking, differential reinforcement of smoking and smoking behavior. The differential peer association scale developed Krohn et al.(1982) and translated by Sohn, Jung-Nam(1999), and the hardiness scale developed by Pollock(1984) and translated by Suh, Mun-Sa(1988) were also used. The data were analyzed using the SAS/PC+ Program and included descriptive statistics, t-test, ANOVA, and Spearman correlation coefficients. The results of this study are as follows : 1. The smoking rate for university students was 50.5% of which 44.7% started smoking in high school. 2. The mean score for level of hardiness was $3.14{\pm}0.43$ 3. The mean score for smoking behavior was as follows : 1) The mean score for the neutralizing definition was $2.16{\pm}0.57$ 2) The mean score for the negative definition was $2.37{\pm}0.71$ 3) The mean score for the positive differential reinforcement was $1.89{\pm}0.63$ 4) The mean score for the negative differential reinforcement was $2.96{\pm}0.64$ 5) The mean score for the differential peer association was $2.67{\pm}1.05$ 4. The data showed positive correlations between hardiness and the neutralizing definition(r=.1951, P<.001), between hardiness and the positive differential reinforcement(r=.1128, P<.05), between hardiness and the amount of smoking per day(r=.1452, P<.05) between the neutralizing definition and positive differential reinforcement(r=.4212, P<.0001), between the neutralizing definition and differential peer association(r=.1856, P<.001), between the neutralizing definition and age at smoking initiation(r=.1582, P<.05), between the negative definition and negative differential reinforcement(r=.2985, P<.0001), between the positive differential reinforcement and differential peer association(r=.3451, P<.0001), between positive differential reinforcement and the amount of smoking per day(r=.4431, P<.0001), between differential peer association and the duration of smoking(r=.2789, P<.0001), between differential peer association and the amount of smoking per day(r=.5410, P<.0001), between the duration of smoking and the amount of smoking per day(r=.4245, P<.0001). The data showed negative correlations between the neutralizing definition and the negative definition(r=-.2065, P<.001) between the neutralizing definition and negative differential reinforcement(r=-.3943, P<.0001) between the neutralizing definition and duration of smoking(r=-.1957, P<.05), between the negative definition and positive differential reinforcement(r=-.2093, P<.001), between the negative definition and the amount of smoking per day(r=-.2282, P<.001), between positive differential reinforcement and negative differential reinforcement(r=-.5555, P<.0001) between negative differential reinforcement and differential peer association(r=-.3653, P<.0001), between negative differential reinforcement and the amount of smoking per day(r=-.4570, P<.0001), between the age at smoking initiation and the duration of smoking(r=-.4594, P<.0001).