Determining the Galactic distribution and numbers of massive stars, such as Wolf-Rayet stars (WRs), is hampered by intervening Galactic or local circumstellar dust obscuration. In order to probe such regions of the Galaxy we can use infrared observat
ions, which provide a means for finding such hidden populations through the dust. The availability of both 2MASS and Spitzer/GLIMPSE large-scale survey data provides infrared colours from 1.25 to 8$mu$m for a large fraction of the inner Galactic plane. In 2005 we initiated a pilot study of the combined set of infrared colours for two GLIMPSE fields and showed that WRs typically occupy a sparsely populated region of the colour space. We followed up 42 of our WR candidates spectroscopically in the near-infrared, and with limited additional observations of some of these candidates in the optical. Six new WRs, four late-type WN and two late-type WC stars, were discovered as a result. Of the remaining $sim$86% of the sample, five appear to be O-type stars. 21 stars are likely of type Be, and 10 stars appear to be of late-type, or possibly young stellar objects, which have contaminated the infrared color space. The survey is generally unbiased towards clusters or field stars, and the new WRs found are in both the field and in and around the RCW 49 region (including cluster Westerlund 2). In this work, and in our other recent work, we show that the infrared broad-band colours to be the most efficient means of identifying (particularly, dust-obscured) candidate massive stars, notably WRs.
