We derive new constraints on the mass of the Milky Way's dark matter halo, based on 2401 rigorously selected blue horizontal-branch halo stars from SDSS DR6. This sample enables construction of the full line-of-sight velocity distribution at different galactocentric radii. To interpret these distributions, we compare them to matched mock observations drawn from two different cosmological galaxy formation simulations designed to resemble the Milky Way. This procedure results in an estimate of the Milky Way's circular velocity curve to image60 kpc, which is found to be slightly falling from the adopted value of 220 km s super(-1) at the Sun's location, and implies image M sub(image). The radial dependence of image, derived in statistically independent bins, is found to be consistent with the expectations from an NFW dark matter halo with the established stellar mass components at its center. If we assume that an NFW halo profile of characteristic concentration holds, we can use the observations to estimate the virial mass of the Milky Way's dark matter halo, image M sub(image), which is lower than many previous estimates. We have checked that the particulars of the cosmological simulations are unlikely to introduce systematics larger than the statistical uncertainties. This estimate implies that nearly 40% of the baryons within the virial radius of the Milky Way's dark matter halo reside in the stellar components of our Galaxy. A value for image of only image M sub(image) also (re)opens the question of whether all of the Milky Way's satellite galaxies are on bound orbits.