The gravity wave drag parametrization of the Whole Atmosphere Community Climate Model (WACCM) has been modified to include the wave‐driven atmospheric vertical mixing caused by propagating, non‐breaking, gravity waves. The strength of this atmospheric mixing is represented in the model via the “effective wave diffusivity” coefficient (Kwave). Using Kwave, a new total dynamical diffusivity (KDyn) is defined. KDyn represents the vertical mixing of the atmosphere by both breaking (dissipating) and vertically propagating (non‐dissipating) gravity waves. Here we show that, when the new diffusivity is used, the downward fluxes of Fe and Na between 80 and 100 km largely increase. Larger meteoric ablation injection rates of these metals (within a factor 2 of measurements) can now be used in WACCM, which produce Na and Fe layers in good agreement with lidar observations. Mesospheric CO2 is also significantly impacted, with the largest CO2 concentration increase occurring between 80 and 90 km, where model‐observations agreement improves. However, in regions where the model overestimates CO2 concentration, the new parametrization exacerbates the model bias. The mesospheric cooling simulated by the new parametrization, while needed, is currently too strong almost everywhere. The summer mesopause in both hemispheres becomes too cold by about 30 K compared to observations, but it shifts upward, partially correcting the WACCM low summer mesopause. Our results highlight the far‐reaching implications and the necessity of representing vertically propagating non‐breaking gravity waves in climate models. This novel method of modeling gravity waves contributes to growing evidence that it is time to move away from dissipative‐only gravity wave parametrizations.