Purpose : Spinal muscular atrophy (SMA) is a degenerative disease of motor neuron caused by loss of the SMN1 gene. SMN2 gene is highly homologous to SMN1 and is known to influence disease severity according to the copy number variation. Our purpose was to investigate genotype-phenotype correlation between the SMN2 copy number and clinical course in SMA patients. Methods : We retrospectively investigated the medical records of SMA patient who received SMN1 and SMN2 mutational analysis by multiple ligation dependent probe amplification in Seoul National University Children’s Hospital from August, 1995 to August, 2012. The clinical subtypes of SMA were classified type I, type II, and type III according to age of onset and severity of the disease. The correlation between SMN2 copy number and clinical severity was analyzed in each clinical subtype. Results : The SMN2 copy number was analyzed in a cohort of 26 SMA patients (male 17, female 9) with homozygous SMN1 gene deletions. The average follow-up duration was 3.7 years (0.2 to 16.7). There were 9 patients with SMA type I, 14 with SMA type II, and 3 with SMA type III. In 9 patients with SMA type 1, the SMN2 copy number was ranged from 2 to 3 (average 2.1). The 8 of 9 patients (88.8%) had two SMN2 copies. There were 4 mortalities in SMA type I group, all of them had 2 SMN2 copies. The 5 patients were on ventilator support, 3 of them had 2 SMN2 copies and 2 of them had 3 SMN2 copies. In 14 SMA type II patients, the SMN2 copy number was ranged from 2 to 4 (average 2.9). The 10 of 14 patients (71.4%) had 3 copies. In 3 SMA type III patients, one patient had 2 SMN2 copies, 2 patients had 3 copies. There was no mortality in type II and III group, and only one patient in type II group who had 3 SMN2 copies was on ventilator support. Conclusion : To our knowledge, this is the first report of SMN2 copy number analysis in Korean SMA patients. The SMN2 gene copies were significantly correlated with clinical subtype of SMA. However, other biomarkers affecting long term clinical outcome including ventilator dependence or mortality need to be further determined in an extended cohort.