본 연구는 한국노동패널조사 4차-12차년도 자료를 바탕으로 실업자 대상 직업훈련의 효과를 검증하였다. 훈련참여자 집단과 비참여자 집단의 구성은 표본선택편의를 최소화하기 위하여 성향점수매칭(PSM)을, 효과 분석은 콕스의 비례위험모형을 이용하여 분석하였다. 실업자 대상의 직업훈련 효과는 ‘재취업확률’, ‘이전임금회복 확률’, ‘재실업확률’을 살펴보았다. 직업훈련의 참여집단의 재취업확률이 비교집단에 비해 매우 유의한 수준으로 높았으나, 사업집단의 이전임금회복확률은 비교집단에 비해 낮았으며, 재실업위험에 있어서는 집단 간 유의한 차이가 없었다. 결과적으로 현재 한국의 직업훈련은 표면적으로 재취업을 촉진하는 역할을 수행하고 있으나, 이는 노동시장으로부터의 이탈을 억제하는 한계적인 역할일 뿐, 임금이나 고용안정성과 같은 고용의 질을 확보하는 기제로서 작동하고 있지는 않음을 확인할 수 있었다. This research verified the effect of job training on the unemployed based on the Korean Labor and Income Panel Study phase 4~12. The subject of this study is 600(540) unemployed people who were dismissed in 2001~2009 and are between the age of 18 to 65. To minimize sample selection bias, PSM method was used while the composition of experimental group and control group and Cox proportional hazard model was used on the analysis. In verifying the effect of job training, the researcher was focused on three rates- ‘re-employment hazard’, ‘the recovery hazard of former income level’, ‘the hazard of recurrence of unemployment after the re-employment’. In conclusion, ‘going through job training’ is surely promoting outplacement in Korea. But it will not guarantee the unemployed the quality of employment, such as income level or employment stability. In other words, job training is leading the unemployed to get employed in short time, by preventing failing from the employe market or falling to the discouraged unemployed. But the quality of employment of them is not significantly higher than the group which were not engaged in job training.