In this study, we have attempted to model the variations in total solar irradiance from two spacecraft. Specifically, we have modeled the Earth Radiation Budget on the Nimbus 7 spacecraft and the active cavity radiometer irradiance monitor (ACRIM-I) on the Solar Maximum Mission (SMM) spacecraft using ground-based photometry of sunspots and faculae from the San Fernando Observatory (SFO). Additionally, for some cases, solar backscatter ultraviolet/2 data on the Mg II core-to-wing ratio from the NOAA 9 spacecraft was used. We have found that most of the solar cycle variation in the total solar irradiance can be accounted for by sunspots and faculae/network. The unexplained variation is not greater than approximately 0.0022% (22 ppm) per year for most of solar cycle 22. Using Nimbus 7 data from March 2, 1985, to December 13, 1993 (1281 data points), as the dependent variable, with the SFO photometric sunspot index (PSI) and the NOAA 9 Mg II core-to-wing ratio for the faculae/network as independent variables (the best model for this interval), we obtained a multiple correlation coefficient squared (R2) of 0.848. The rms noise in the residuals is approximately 0.221 W m-2 (162 ppm). This rms noise appears to be dominated by noise in the spacecraft data. For the same model, but for the time interval from March 2, 1985, to July 14, 1989, we obtained an R2 of 0.838 for 718 data points. The same type of model for this same interval, substituting SMM/ACRIM-I total irradiance for Nimbus 7, gave an R2 of 0.857 for 685 data points. Our best correlation, however, came from a three-parameter model, fitting Nimbus 7 data to the SFO digital PSI, the SFO facular index PFIFA, and the NOAA 9 data for the interval May 30, 1988, to December 13, 1993, giving an R2 of 0.887 (745 data points). These strong correlations suggest that most of the variation in solar irradiance is associated with known solar magnetic features. Whether or not these magnetic features can explain all of the solar irradiance variability will require more stable and accurate long-term measurements from space and the ground. ¿ American Geophysical Union 1996