<p>This paper presents a large-signal analysis of radiofrequency power amplifier (PA) efficiency under the constraint of conjugate output matching. The study shows that in intrinsically low-efficiency operating regimes such as class-A, enforcing conjugate output matching leads to a factor-of-two reduction in efficiency with respect to an equivalent unmatched design, in agreement with the literature. However, when the amplifier is operated in class-AB and beyond, the efficiency penalty is reduced. Consequently, by appropriately adjusting the drive level and the PA bias point, it is possible to realize an output-matched PA with efficiencies comparable to those of unmatched designs. In addition to the analytical treatment, a practical topology and a design methodology for achieving the required output matching are presented. The findings are validated through transistor-level simulations of both bipolar and MOS RF amplifier designs.</p>