During mitosis, genomic integrity is maintained by the proper coordination of anaphase entry and mitotic exit via mitotic checkpoints. In budding yeast, mitotic exit is controlled by a regulatory cascade called the mitotic exit network (MEN). The MEN is regulated by a small GTPase, Tem1p, which in turn is controlled by a two-component GAP, Bfa1p-Bub2p. Recent results suggested that phosphorylation of Bfa1p by the polorelated kinase Cdc5p is also required for triggering mitotic exit, since it decreases the GAP activity of Bfa1p-Bub2p. However, the dispensability of GEF Lte1p for mitotic exit has raised questions about regulation of the MEN by the GTPase activity of Tem1p. We isolated a Bfa1p mutant, $Bfa1p^{E438K}$, whose overexpression only partially induced anaphase arrest. The molecular and biochemical functions of $Bfa1p^{E438K}$ are similar to those of wild type Bfa1p, except for decreased GAP activity. Interestingly, in $BFA1^{E438K}$ cells, the MEN could be regulated with nearly wild type kinetics at physiological temperature, as well as in response to various checkpoint-activating signals, but the cells were more sensitive to spindle damage than wild type. These results suggest that the GAP activity of Bfa1p-Bub2p is responsible for the mitotic arrest caused by spindle damage and Bfa1p overproduction. In addition, the viability of cdc5-2 ${\Delta}bfa1 $ cells was not reduced by $BFA1^{E438K}$, suggesting that Cdc5p also regulates Bfa1p to activate mitotic exit by other mechanism(s), besides phosphorylation.