A radiation-driven disk wind model is proposed that offers great promise of explaining the extreme mass loss signatures of massive young stellar objects (the BN-type objects and more luminous Herbig Be stars). It is argued that the dense low-velocity winds associated with young late-O/early-B stars would be the consequence of continuing optically-thick accretion onto them. The launch of outflow from a Keplerian disk allows wind speeds of ~200 km s−1 that are substantially less than the escape speed from the stellar surface. The star itself is not required to be a rapid rotator. Disk irradiation is taken into account in the hydrodynamical calculation presented, and identified as an important issue both observationally and from the dynamical point of view.