Comments on Duality in MQCD

We clarify some ambiguous points in a derivation of duality via brane exchange using M-theory language, and propose a ``proof'' of duality in MQCD. Actually, duality in MQCD is rather trivial and does not need a complicated proof. The problem is how to interpret it in field theory language. We examine BPS states in N=2 theory and find the particle correspondence under duality. In the process, we also find some exotic particles in N=2 MQCD, and we observe an interesting phenomenon in type IIA string theory, namely, that fundamental strings are converted into D2-branes via the exchange of two NS5-branes. We also discuss how we should understand Seiberg's N=1 duality from exact duality in MQCD.