A QCD chiral critical point at small chemical potential: is it there or not?

For a QCD chiral critical point to exist, the parameter region of small quark masses for which the finite temperature transition is first-order must expand when the chemical potential is turned on. This can be tested by a Taylor expansion of the critical surface (m_{u,d},m_s)_c(mu). We present a new method to perform this Taylor expansion numerically, which we first test on an effective model of QCD with static, dense quarks. We then present the results for QCD with 3 degenerate flavors. For a lattice with N_t=4 time-slices, the first-order region shrinks as the chemical potential is turned on. This implies that, for physical quark masses, the analytic crossover which occurs at mu=0 between the hadronic and the plasma regimes remains crossover in the mu-region where a Taylor expansion is reliable, i.e. mu less than or similar to T. We present preliminary results from finer lattices indicating that this situation persists, as does the discrepancy between the curvature of T_c(m_c(mu=0),mu) and the experimentally observed freeze-out curve.