Title: A robust estimate of the Milky Way mass from rotation curve data

Abstract: We present a new estimate of the mass of the Milky Way, inferred via a Bayesian approach by making use of tracers of the circular velocity in the disk plane and stars in the stellar halo, as from the publicly available {\tt galkin} compilation. We use the rotation curve method to determine the dark matter distribution and total mass under different assumptions for the dark matter profile, while the total stellar mass is constrained by surface stellar density and microlensing measurements. We also include uncertainties on the baryonic morphology via Bayesian model averaging, thus converting a potential source of systematic error into a more manageable statistical uncertainty. We evaluate the robustness of our result against various possible systematics, including rotation curve data selection, uncertainty on the Sun's velocity $V_0$, dependence on the dark matter profile assumptions, and choice of priors. We find the Milky Way's dark matter virial mass to be $\log_{10}M_{200}^{\rm DM}/ {\rm M_\odot} = 11.92^{+0.06}_{-0.05}{\rm(stat)}\pm{0.28}\pm0.27{\rm(syst)}$ ($M_{200}^{\rm DM}=8.3^{+1.2}_{-0.9}{\rm(stat)}\times10^{11}\,{\rm M_\odot}$). We also apply our framework to Gaia DR2 rotation curve data and find good statistical agreement with the above results.