Modeling studies are carried out using the Sheffield University plasmasphere-ionosphere model to investigate the relative importance of neutrals winds, neutral densities, and solar EUV fluxes in leading to the saturation of ionospheric ionization observed during the intense solar cycle 21. Values of mean daytime (1100--1700 LT) ionospheric electron content, peak electron density (Nmax), and peak height (hmax) computed for four midlatitude stations for the month of October 1980--1985, when the 10.7-cm solar activity index (F10.7) varied from 66 to 303, increase nonlinearly with F10.7 with saturation for hig values of F10.7. Neutral winds, neutral densities, and solar EUV fluxes (obtained from empirical models based on observed data) used in the model computations also undergo similar nonlinear increase with F10.7. The study reveals that (1) the nonlinear increase of neutral winds and neutral densities has no net effect on the saturation of ionospheric ionization, and (2) the saturation of ionization is caused by the saturated production of ionization due to the nonlinear increase of the solar EUV fluxes. The model values of ionospheric height saturate mainly due to the nonlinear increase of the neutral densities determined by solar EUV and UV radiations. The study concludes that the ionosphere (and atmosphere) responds linearly to the solar EUV (and UV) inputs and nonlinearly to F10.7 because the expected linear relationship between the EUV (and UV) fluxes and F10.7 breaks down during solar maximum. Thus it is recommended that the F10.7 proxy, which has conventionally been used as an index of solar activity, must be replaced or be used carefully, particularly during solar maximum. ¿ American Geophysical Union 1994