We build a sample of distant (D > 80 kpc) stellar halo stars with measured radial velocities. Faint (20 < g < 22) candidate blue horizontal branch (BHB) stars were selected using the deep, but wide, multi-epoch Sloan Digital Sky Survey photometry. Follow-up spectroscopy for these A-type stars was performed using the Very Large Telescope (VLT) FOcal Reducer and low dispersion Spectrograph 2 (FORS2) instrument. We classify stars according to their Balmer line profiles, and find that seven are bona fide BHB stars and 31 are blue stragglers (BS). Owing to the magnitude range of our sample, even the intrinsically fainter BS stars can reach out to D ∼ 90 kpc. We complement this sample of A-type stars with intrinsically brighter, intermediate-age, asymptotic giant branch stars. A set of four distant cool carbon stars is compiled from the literature and we perform spectroscopic follow-up on a further four N-type carbon stars using the William Herschel Telescope (WHT) Intermediate dispersion Spectrograph and Imaging System (ISIS) instrument. Altogether, this provides us with the largest sample to date of individual star tracers out to r ∼ 150 kpc. We find that the radial velocity dispersion of these tracers falls rapidly at large distances and is surprisingly cold (σr ≈ 50–60 km s−1) between 100 and 150 kpc. Relating the measured radial velocities to the mass of the Milky Way requires knowledge of the (unknown) tracer density profile and anisotropy at these distances. Nonetheless, by assuming the stellar halo stars between 50 and 150 kpc have a moderate density fall-off (with power-law slope α < 5) and are on radial orbits (), we infer that the mass within 150 kpc is less than 1012 M⊙ and suggest it probably lies in the range (5–10) × 1011 M⊙. We discuss the implications of such a low mass for the Milky Way.