Increasing quantities of and access to both wildlife survey data and non-designed incidental or citizen science data have left us with a rather big problem: how to we put all of these disparate pieces together and build sp ecies distribution models that use as much of the available data as possible? This leads us to a series of sub-questions that I will address in this talk: should we combine data then model it all at once or, build multiple models and gure out how to combine their outputs (or couple their fitting)? How can we find equivalences in recorded effort (and what can we do when no effort is recorded)? I’ll illustrate these issues and offer some solutions using example data from aerial and shipboard surveys of seabirds in New England, as well as from large-scale surveys of marine mammals in the North Atlantic.