This study has been conducted on the nonequivalent control group pretest-posttest design in quasi experimental basis and newly born premature infants from intensive care unit of G Medical University Hospital in Inchon Metropolitan were selected in two groups of 21 infants each. The first group for experimental and the other for control. Data has been collected form October 30, 1997 to August 29, 1998. For the experimental group tactile and kinesthetic stimulation was applied 2 times a day for 10 days(10 : 00~11 : 00 hours in the morning and 17 : 00~18 : 00 in the afternoon). As a weight weighing instrument. electronic indicator scale(Cas Co. korea) was used. To determine urine cortisol concentration level in stress hormone, radio immune assay method was used. And high performance liquid chlomatography was used to determine urine norepinephrine, concentration level To determine behavior status, tools developed by Anderson et at(1990) and remodeled by Kim Hee-Sook(1996) were used. Collected data were analyzed with the SAS program using x$^2$-test, student t-test, repeated measures ANOVA and paired t -test. The result were as follow. 1. As for the daily weight gain. the experimental group showed first change in weight and this group also showed higher weight in the average weight than the control group. Statistically, however, there was no significant factor between the two group. 2. The cortisol concentration in urine showed decrease in the experimental group norepinephrine concentration in urine showed increase in both experimental and control groups. No statistical significance was shown between the two groups. 3. In the aspect of behavior status. the experimental group showed statistical significance by showing inactive in the state of alert and conversion to a positive state than the control group. In conclusion, the sensory stimulation in this study showed a positive aspect through there was no statistical significance in the weight gain and urine stress hormone concentration. In the behavior status, there was statistical significance in the frequency of staying inactive in the state of alert and conversion to a positive state.