Despite a broad consensus in policy circles and scholarship that NATO member states free ride, credible causal evidence remains scarce. We assemble a matched panel of post-1990 entrants and comparable non-members and estimate the impact of NATO membership using modern panel difference-in-differences methodology that accommodates staggered adoption and heterogeneous treatment. To address anticipation, we analyze not only formal accession but also the integration step of the NATO Membership Action Plan (MAP) as a treatment and complement our analysis by qualitative evidence from archival (declassified) and official documents. Across specifications, we find no statistically significant effect of either accession or MAP participation on military expenditure as a share of GDP. Yet we go beyond reporting null results by deriving one-sided bounds that rule out a significant medium- and longer-term average free-riding effect of NATO membership. Ultimately, by itself, joining NATO does not seem to induce meaningful cutbacks in military spending. Accession appears more consistent with strengthening security than with creating opportunities to free ride.