This paper re-examines the validity of the Phillips-Curve framework using US data. We make three main innovations. First, we introduce into the well-known Calvo price staggering framework, a regime-dependent price-changing signal. This means that a state-dependent linearization is no longer required to derive the Phillips relationship and thus that questions of regime dependency can be addressed. Second, we engage on a careful modeling of long-run supply in the economy, which permits more data-coherent measures of output gaps and real marginal costs indicators consistent with underlying, frictionless supply. Finally, we include two types of labor adjustment costs reflecting the intensive and extensive participation decisions. As regards the latter, we introduce the concept of ?effective? working hours into the production technology which generates an overtime function directly into the mark-up equation. This, it turns, out has first-order implications for the cyclicality and econometric fit of the mark-up implied by the Phillips-curve representation.