This chapter examines the use of reference and analysis in the study of astronomical systems from Newton to the present day. I begin by noting that Bas van Fraassen pays more attention to theoretical models that do the work of analysis, while Nancy Cartwright and Margaret Morrisson pay more attention to models where the relevant relation is not satisfaction (as between a meta-language and an object-language), but representation (as between a discursive entity and a thing that exists independent of discourse). I then track the use of both kinds of models in Newton’s Principia, Books I and III, and find strategies of juxtaposition, superposition, and unification. In the era after Newton, problems of analysis were addressed by Euler, Lagrange, Laplace and Hamilton, while problems of reference came to the fore in the empirical work of Herschel and Rosse. The tension between reference and analysis also appears in the debates between Hubble and Zwicky, and in the work of Vera Rubin.