An integrated electrical–optical model has been used to analyze the performance of hydrogenated amorphous silicon (a-Si:H) p-i-n solar cells having a p-type hydrogenated microcrystalline silicon (p-μc-Si:H) window layer. Our attempts to simulate various experimentally measured solar cell characteristics of such cells available in the literature indicate that for the same thickness, μc-Si:H p layers may have different mobility band gaps (Eμ), which may be linked to different crystalline volume fractions (Fc). Modeling reveals that there is both an optimum value for Eμ (therefore Fc), as well as an optimum thickness of the μc-Si:H p layer for obtaining the best solar cell performance. A thin (8–10 nm) film having a rather low crystalline volume fraction (high Eμ) was found from our computer analysis to give the best results. This is because as the band gap of μc-Si:H p layer decreases (i.e., its crystallinity increases), the valence band offset at the p/i interface increases, leading to a collapse of the bulk electric field. Also investigated was the effect of sandwiching an intrinsic or boron-doped carbide buffer layer between the μc-Si:H p layer and the intrinsic absorber. In both cases the open-circuit voltage (Voc) and the fill factor (FF) improve markedly on account of the reduction of the large potential drop at the p/i interface, which ensures more field over the intrinsic absorber layer. However, whereas for an i-a-SiC:H buffer- , the cell performance still depended on the mobility band gap of the μc-Si:H p layer; for a p-a-SiC:H buffer, the conversion efficiency (η) could be made to attain nearly the same value, regardless of Eμ(p-μc-Si:H), simply by adjusting the thickness of the p-a-SiC:H layer. Compared to a “standard” p-i-n cell employing a single p-a-SiC:H window, both combinations show improvements in Voc and FF, but a fall in the short-circuit current density (Jsc), leading to a small net increase in η. © 1999 American Institute of Physics.