|Abstract This dissertation investigates massive star Formation in the Milky Way via millimeter and submillimeter observations of luminous young stellar objects. First, I describe a snapshot continuum and multi-molecular-line survey with the BIMA array aimed at characterizing very young high-mass stars. The target sample consists of eleven luminous IRAS sources associated with very young ultracompact (UC) HII regions, which have weak cm-wave flux densities and very compact angular sizes despite their small (kinematic) distances. In this way, I hope to preferentially select and study extremely young high-mass stars. Most of the 3 millimeter continuum emission shows multiple components, which is consistent with the picture of a clustered formation mode of massive stars. I identify five 3 mm dust sources as good candidates for high-mass protostars because they are coincident in position with their luminous IRAS counterparts and dense molecular gas as traced by H13CO+ emission, yet not detected at 3.6 cm. The typical dust and gas mass of the 3 mm components is a few tens M_sun, while the 3 mm components with centimeter counterparts are ~5 times less massive than those not associated with centimeter continuum, suggesting that some 3 mm sources with centimeter counterparts may relate to very young intermediate-mass stars. On the other hand, due to the relatively weak centimeter luminosity of a few mJy kpc^2, some target centimeter continuum sources could correspond to stellar winds/jets rather than real HII . Furthermore, I present near-IR photometry of the target compact centimeter sources obtained from the new released Two Micron All Sky Survey archival data. Second, I present millimeter observations with the BIMA array of the bipolar molecular outflows associated with the luminous far-IR sources IRAS 21519+5613 and IRAS 22506+5944. Although outflows have been identified as a common occurrence in the formation of both high-and low-mass stars, only about ten molecular outflows associated with luminous young stellar objects have been studied in high angular resolutions (<10"). I find that the outflow associated with each IRAS source shows a clear bipolar morphology in 12CO, with properties (i.e., total mass of order 10-100 M_sun, mass-outflow rate >10^-3 M_sun yr^-1, dynamical timescale 10^4-10^5 yrs, and energetics) comparable with those of other massive outflows associated with luminous young stellar objects. Each outflow appears to be centered on a dust and gas condensation with a mass of 200-300 M_sun, likely marking the location of the driving source. The mass-velocity diagrams of both outflows change in slope at a velocity of ~10 km s^-1, suggesting that the high-velocity component (HVC) may drive the low-velocity component (LVC). Neither HVC has a momentum supply rate sufficient to drive their corresponding LVCs, and as for other molecular outflows the primary driving agent cannot be ionized gas, leaving atomic gas as the other remaining candidate. Finally, I present a first glimpse at the small-scale distribution of dust around UC HII regions and high-mass protostellar objects (HMPOs) at 0.85 mm made with the Submillimeter Array (SMA), and investigate whether HMPOs and UC HII regions can serve as useful gain calibrators for the SMA. The SMA observations with an angular resolution of ~3" reveal that although a large fraction of the dust emission originates from an extended ``halo' component, a compact unresolved component often remains that when sufficiently strong may serve as a useful calibrator. I discuss the origin of the core-halo structure seen in many sources and conclude with suggestions for future searches for calibrators with the SMA.