Line-commutated converters (LCCs) are widely used in various high-power applications such as generator-rectifier systems, exciters, front-end rectifier loads, and classic high voltage dc systems. Among various techniques used for modeling LCC systems, the dynamic average value modeling (AVM) wherein the effect of switching is neglected or averaged over a prototypical switching interval has become indispensible since it results in continuous, linearizable, and computationally efficient models. The conventional AVMs can only predict the fundamental component of the ac voltages and currents, and neglect the harmonics injected at the ac side by the switching converter. In this paper, a recently proposed parametric average-value modeling (PAVM) approach is extended using multiple reference frame theory to include the significant harmonics of interest (e.g., fifth and seventh) for diode rectifiers. The new PAVM is verified against the detailed simulation in steady-state and transient studies, and is effective in predicting the transient waveforms, while achieving significant computational advantage (speed up) in time-domain simulation over conventional models.