The effect of long acting bronchodilators (LA-B) is well documented in various randomized controlled studies. However, it is not evident that this effect is still consistent in real practice. In this study, we analyzed the effect of LA-B for newly-diagnosed COPD patients. We analyzed Medicaid data from 2007 to 2010. Newly-diagnosed COPD patients were defined if the participants were ≥ 40 years old had never been registered as ``COPD`` (J42-44) for the past 2 years, then registered as COPD and prescribed COPD medication in 2009. LA-B group was defined if LA-B was prescribed ≥2 times in 2009. Short acting bronchodilator (SA-B) group was defined if LA-B was prescribed ＜2 times or only SA-B was prescribed. Oral medication (OM) group was defined if bronchodilator was never prescribed in 2009. A total of 77,800 patients were enrolled. 8,663 (71.5%) was male. Mean age was 66.0±10.5 years. LA-B group included 12,115, SA-B 5,173, and OM 60,192 patients. ER visit and hospitalization was associated with SA-B group, male, older age, Medicaid aid, Tertiary hospital utilization and more co-morbidity. ER visit, recurrent ER visit, hospitalization and recurrent hospitalization were increased in SA-B group in multivariate analysis (adjusted OR [95% confidence interval]=4.32 [3.93-4.75], 6.19 [5.24-7.30], 5.04 [2.95-3.39], 8.49 [7.67-9.39]) and it resulted in high medical cost. Long acting bronchodilator in initial medication showed lower hospitalization, ER visit and medical cost for newly-diagnosed COPD patients in real practice.