There have been several attempts in developing countries to reduce both graduate and overall unemployment; with the majority attempts centered on changes in the education sector. To better understand this avenue, this study intends to comparatively establish the impact of the two broad discipline categories of- Arts and science related disciplines- on the overall unemployment. The study employed the Vector autoregressive (VAR) model to analyse Uganda's data between 1991 and 2017. The findings of the study showed that the arts/humanities graduates have a slightly higher impact on unemployment than their counterparts from the science/technology disciplines in the short run in Uganda; with both groups of graduates having no significant effect on unemployment in the long run.