We investigate the effects of polydispersity on dielectric and viscoelastic normal-mode relaxation of cis-polyisoprene melts above entanglement. Measured dielectric and mechanical spectra and relaxation moduli of polydisperse samples are compared to the predictions of models combining either simple reptation, double reptation or des Cloizeaux’ time-dependent-diffusion single-chain autocorrelation functions (ACF) to linear or nonlinear (Tsenoglou) mixing rules. The blending rules are tested by employing experimental data, thereby eliminating uncertainties associated with the experimental molecular weight distribution (MWD) and choice of specific forms for the ACFs. Predictions from the various models based on the MWD from size exclusion chromatography are compared to the experimental data as well. We find that dielectric spectra are satisfactorily captured by a Linear mixing rule, but quantitative description of the theology requires the use of the nonlinear (Tsenoglou) mixing rule. The single-chain ACF of reptation with a 3.7 power-law dependence of the relaxation time on molecular weight accurately predicts both dielectric and viscoelastic spectra. The double reptation ACE while being inadequate in the dielectric case, gives viscoelastic predictions that are nearly identical to those obtained with the simple reptation ACE The des Cloizeaux ACF fails in the dielectric case and is less satisfactory than simple or double reptation in the viscoelastic case. This limited evidence indicates that the ACFs for end-to-end vector reorientation (dielectric relaxation) and entanglement "dissolution" (viscoelastic relaxation), although both originating in chain reptation, reflect different mechanisms.