The purposes of this research were to add to the developing knowledge base about the burden and social support of grandmothers involved with caring grandchildren, and to examine whether a significant relationship exists between their burden and social support perceived by the grandmothers. A convenience sample of 102 grandmothers was recruited from five collective apartment areas at a small city in Korea. The criteria of selection of sample were that the grandmothers were raising their grandchildren under the age of 36 months for 3 months or longer at own or gandchildren's home. The instruments used were a 15-item multidemensional burden scale and a 8-item social support scale. Both scales were self report, five point Likert type scales. The higher the score, the higher the degree of burden and social support. Data was collected by two prepared research assistants visiting subjects' home from December 10, 1998 to March 20, 1999. The collected data were analysed using mean, t-test, one-way ANOVA, and Pearson's correlation coefficient computed by SPSS software. The results were as follows. 1. In the age distribution of grandmothers, the over half of subjects(58.8%) were under 60 years old. The majority(69.6%) of subjects were married at time of data collection. 79(77.5%) of the sample reported that their perceived health status was good or over. 2.The burden of the grandmothers was not scored high, and the item means on burden scale were ranged from 2.26 to 4.19 out of 5. 'Short of private time' (4.19) had the highest score, followed by 'fatigue' (3.92), 'short of rest' (3.75), and 'short of contact with friends and neighbors' (3.62). The lowest item was 'family doesn' t help me' (2.26), and followed by 'family doesn' t understand me' (2.33), 'angry with family' (2.43), and 'angry while caring for grandchild' (2.60). 3. There were significant relationships between the burden and present health status(p < .01), childcaring confidence(p < .01), and motive of caring(p < .01). 4. The score of social support, was ranging from 3.61 to 4.01 out of 5. 5. The relationship between burden and social support was found to be correlated negatively. The relationship was statistically significant( gamma = - .2833, p < .001). In conclusion, it was found that the burden was not high and burden of grandmothers caring grandchildren was correlated negatively to social support. Therefore, these results provide a basis for developing a nursing intervention to reduce the burden of grandmothers.