The prediction of polymer/polymer miscibility is addressed using analog calorimetry and molecular modeling. For each polymer, an analog compound representing one or two repeat units was chosen. Heat-of-mixing was measured for liquid mixtures of analog compounds and then used in a binary interaction model to predict polymer miscibility. Specifically, we have measured exothermic heats-of-mixing for 4-ethyl phenol, an analog of poly(vinyl phenol), with several analogs containing ether, ester, or ketone functional groups. The exothermic heat-of-mixing results are consistent with the observed miscibility of poly(vinyl phenol) with polymers containing these functional groups. Using interaction parameters derived from the analog calorimetry in the binary interaction model or using premixes of 4-ethyl phenol in ethyl benzene, we correctly predict the magnitude and relative order of the fraction of vinyl phenol units in copolymers with styrene required for miscibility with poly(methyl methacrylate), polyacetal, and a polyketone. The miscibility trends for poly(vinyl phenol) blends predicted from analog calorimetry and the binary interaction model are in reasonable agreement with those predicted from the association model of Painter and Coleman, despite the different bases of the two approaches. We have used molecular modeling to complement the analog calorimetry and to assess steric effects on hydrogen-bonding ability for models of poly(n-butyl acrylate) and poly(t-butyl acrylate) with phenol. The modeling results suggest that, in some cases, steric effects and the three-dimensional structure of the polymer can significantly influence the hydrogen-bonding ability of polymers relative to their analogs.