Ground-based near-IR high resolution spectra of the H I lines in luminous YSOs have revealed that many of these objects have in addition to the broad optically thick component from an ionized outflow, a narrow optically thin component possibly originating where the wind interacts with other circumstellar material. In an object like LkHa101 this component dominates to the extent that we cannot be sure there is a wind. It is not clear whether this is an evolutionary trend or is more to do with the eventual main sequence mass of the star. We can unravel the hybrid nature of the H I lines by going to higher delta n transitions up a series. The decrements up a series are such that even where the optically thin component dominates initially, an optically thick component must take over at some point, if present. Efforts to investigate this from the ground with low n series are severely hampered by large and uncertain reddening corrections. We propose to observe numerous members of the n=6,7, and 8 series of H I in the 4.1-7.5 micron range where the extinction is low. Scanning this range will no doubt also yield many other previously unseen lines. The capabilities of the SWS also allow us to simultaneously obtain a spectrum of the H I 10-8 16.2 micron line at high spectral resolution with the FP with SNR sufficient to define the line profile. This high level line will give us valuable information on the velocity field in the innermost circumstellar medium (CSM), close to the star. Such data are vital to understanding the origin of the ionized outflows characteristic of this object class. We also propose to observe in the same manner two post-main sequence B[e] stars (a class known to exhibit strikingly similar line emission). Like our Herbig Be target, MWC 297, these stars are optically-accessible in a way that M8E and S106IR are not. Comparisons among their SWS spectra will act as a bridge to furthering our understanding of the nature of the outflows in luminous YSOs.