Snowfall is an important weather element with substantial societal impacts. Snow climatology is likely to be highly sensitive to possible anthropogenic influences on the climate system. It is also a key component of the climate system because of the feedback effects arising from its high albedo. The Climate Data Modernization Program has recently keyed pre-1948 daily snow observations, providing the opportunity to extend trends analyses back to the beginning of the 20th Century. Initial analyses of these data indicated sizeable and intriguing trends. These included downward trends at low snowfall stations and upward trends at high snowfall stations. These features could reflect net northward shifts in the tracks of snow-producing systems and greater water vapor availability, both changes perhaps suggestive of anthropogenic influences. However, our investigations also revealed numerous apparent inconsistencies in the snowfall data, casting suspicion on the reality of the above trends. For example, long-term snowfall trends among neighboring stations differ greatly in some cases, a feature that is unlikely to be a real physical phenomenon. Internal inconsistencies in the snow records, such as lack of upward trends in maximum seasonal snow depth at stations with large upward trends in snowfall, point to inhomogeneities. Thus, great care must be taken in interpretation of time series of snow-related variables. To address this, we have undertaken an intensive project to examine all of the long-term snowfall stations and are identifying a set of stations we believe to be suitably homogeneous for trends assessments. We will present long-term snowfall trends for this homogeneous station set.