In this reply, we address the comment by Ericsson and Sjoqvist on our paper [Phys. Rev. A {bf 84}, 034103 (2011)]. We point out that the zero gauge field is not the evidence of trivial geometric phase for a non-Abelian SU(2) gauge field. Furthermore, the recalculation shows that the non-Abelian geometric phase we proposed in the three-level $Lambda$ system is indeed experimentally detectable.