Decentralization is expected to reduce poverty. Through fiscal decentralization, poverty may be reduced by optimizing intergovernmental grants. However, its effect on poverty is beginning to be questioned after some districts in Indonesia showed slowing performance on poverty reduction despite increasing transfer. Thus, this paper tends to investigate the linkage of intergovernmental transfer (IT) and poverty in Indonesia. By using panel data from 2005 to 2013, the linkage is tested using linear regression model. The result shows that increasing 1% of IT allocation will reduce poverty absolute by 0.12%, or 10% increase of IT will reduce poverty by 1.24%. The intergovernmental transfer also directly affects poverty gap and poverty rate. Increasing 10% of IT will reduce the poverty gap and poverty rate by 9.49% and 31.73%, respectively. Moreover, DAU (unconditional intergovernmental transfer) has a higher effect than DAK (conditional intergovernmental transfer) on poverty eradication. An increase of DAU by 1% will reduce the poverty number by 0.068%, in contrast, DAK only has 0.0418% to reduce poverty. However, as the difference is relatively small, the effectiveness of DAU on poverty is not consistent with the decentralization theory. Unconditional grant must have a significant difference than conditional grant on poverty. It may indicate that the DAU has been spent ineffectively and inefficiently. It may also indicate that it has been spent on not pro-poor policies. Thus, fiscal decentralization in Indonesia needs to be reevaluated in order to support poverty alleviation program.