Testing m(u)=0 on the lattice

A massless up quark is an intriguing possible solution to the strong CP pro blem. We discuss how lattice computations can be used in conjunction with c hiral perturbation theory to address the consistency of m(u) = 0 with the o bserved hadron spectrum and interactions. It is not necessary to simulate v ery light quarks - three flavor partially quenched computations with compar able sea and valence quark masses on the order of the strange quark mass co uld suffice.