The field line interhemispheric plasma (FLIP) model is used to study the 6300 ¿ line intensity measured during three morning twilights from the McDonald Observatory in Texas. The Imaging Spectrometric Observatory (ISO) measured the 6300 ¿ intensity during the winter of 1987 and the spring and summer of 1988. The FLIP model reproduces the measured intensity and its variation through the twilight well on each day using neutral densities from the MSIS-86 empirical model. This is in spite of the fact that different component sources dominate the integrated volume emission rate on each of the days analyzed. The sensitivity of the intensity to neutral composition is computed by varying the N2, O2, and O densities in the FLIP model and comparing to the intensity computed with the unmodified MSIS-86 densities. The ion densities change self-consistently. Thus the change in neutral composition also changes the electron density. The F2 peak height is unchanged in the model runs for a given day. The intensity changes near 100¿ SZA are comparable to within 10% when either [O2>, [N2>, or [O> is changed, regardless of which component source is dominant. There is strong sensitivity to changes in [N2> when dissociative recombination is dominant, virtually no change in the nighttime (SZA≥108¿) intensity with [O2> doubled, and sensitivity of over 50% to doubling or halving [O> at night. When excitation by conjugate photoelectrons is the dominant nighttime component source, the relative intensity change with [O> doubled or halved is very small. This study shows the strong need for simultaneous measurements of electron density and of emissions proportional to photoelectron fluxes if the 6300 ¿ twilight airglow is to be used to retrieve neutral densities. ¿ American Geophysical Union 1995