The abundance ratio of D:H in the interstellar medium can be derived either more directly from the UV absorption measurements of HD and H2 in the spectra of early-type stars, or less directly from millimeter-wave emission in the rotational lines of a variety of molecules. The latter method is applicable over a much larger portion of the galaxy, but is subject to the uncertainty inherent in chemical fractionation. It is shown here how the basic assumptions of temperature distribution and geometry within a single source can cause the derived ratio HDO:H2O to vary between 0.05 and 0.0025, a comparable range to that found with three different molecules DCN:HCN, DNC:HNC, and DCO(+):HCO(+) in the same source. This result should lead to caution in the interpretation of such abundance ratios and of their gradients within the galaxy.