ABSTRACT Stellar evolutionary isochrones are coupled with mostly theoretical flux distributions of stars in order to predict the flux distributions and colors of stellar populations as a function of population age and metallicity. Twenty one empirical indices of absorption feature strengths are measured for ensembles of stars of differing ages, temperatures, and metallicities. These indices are expressed as polynomials in stellar temperature, surface gravity, and metallicity, and are applied to the population models to predict index strengths as a function of population age and metallicity. Checks are made against stellar observations. The model predictions are compared to measured indices of elliptical galaxies. The logarithmic ratio [Me/Fe] (which is zero for a scaled solar ratio) is shown to be greater than that of the most metal-rich stars in the solar neighborhood in giant elliptical galaxies. Since [Mg/Fe] is probably near zero for metal-rich stars, [Mg/Fe] appears to be greater than zero in giant ellipticals, but not in smaller, M32-class, ellipticals. This behavior is probably mirrored in Na and NC indices, suggesting that all light elements are relatively enhanced in giant ellipticals, on average. If this is true, the nucleosynthetic history of giant ellipticals must differ from that of smaller ellipticals. Surface brightness fluctuation magnitudes are predicted. Models agree with the empirical I-band calibration, giving extragalactic distances which may be good to 10% for galaxies well-measured in two passbands. Qualitative agreement is also reached in fluctuation color-color diagrams and color-index diagrams. A fluctuation magnitude which is independent of color probably exists between Cousins I and Johnson J. The model colors, magnitudes, mass-to-light ratios, spectral indices, and fluctuation magnitudes are tabulated. Four specific theorems follow from this work. First, the stellar populations of ellipticals are controlled by at least three parameters, and these parameters are most likely mean heavy element abundance, mean light element abundance, and mean age. Second, if Delta-age/Delta-Z = 3/2 for two populations, they will appear almost identical in most indices. The few exceptional indices include H-Beta and the G band, which provide the best age sensitivity, and several Fe blends which are roughly twice as sensitive to metallicity as any previously used index assuming the absence of abundance ratio differences. Third, of the controlling parameters of a population, Z is dominant, followed by age, IMF, and Y. For the present, uncertainties in stellar evolution have the same magnitude as the effects of IMF and Y on population indices. Finally, because of the similar effects of age and metallicity on absorption features, the spectra of old populations are sensitive to changes in some elemental abundance ratios (like [Mg/Fe]), implying that these ratios can be quantitatively measured when appropriate models become available.