Reports of party membership have documented steep declines in formal enrollment in many countries, but numbers alone tell us little about the systemic causes or likely implications of such drops. Existing cross-national surveys can help illuminate the extent and consequences of these changes. Unfortunately, they are difficult to use for this purpose because differences in question wording and institutional differences in the meaning of party membership produce unstable results. This paper proposes to overcome these obstacles by using behavioral measures of party membership in survey-based studies of party membership. We demonstrate that such measures produce results which are more consistent across surveys and therefore more meaningful. We then illustrate their utility by using a behaviorally-qualified measure to study how party membership decline has affected parties’ ability to mobilize grassroots partisan participation. Our analysis also shows that even though reported party membership declined from the 1990s to the 2000s, there was surprisingly little change in this decade in the number of politically active partisans.