Photometry has been obtained in the V, I (Cousins) system for a large number of giants in six Galactic globular clusters that cover a wide range of metal abundance. The data were derived from CCD observations on the 0.9 m telescope at CTIO over three successive seasons. Both internal checks, and external comparisons with globular cluster stars measured by conventional photoelectric photometry, suggest that the CCD photometry is accurately on the standard system, and that the cluster star colors and magnitudes have overall uncertainties of less than 0.02 mag. These data, together with a cluster distance scale, which we take as that derived from recent horizontal branch models, allow the construction of cluster giant branches in a (MI, (V - I)0) color-magnitude diagram. This diagram is then used to derive metal abundance estimates, relative to the abundances assumed for the calibrating clusters, for the globular clusters Pal 12 and Eridanus from V, I photometry of their giants. The abundances derived are [Fe/H] = - 1.06 ± 0.12 and [Fe/H] = - 1.50 ± 0.15, respectively. Further, the photometry is used to set 3σ upper limits in the range 0.04-0.09 dex for any intrinsic heavy-element abundance dispersion in the six calibrating clusters. The observed giant branches are also compared with the predictions of theory. In particular, we give a calibration of the bolometric correction to I magnitudes as a function of (V - I)0 color. With this calibration and our adopted distance scale, we compare the bolometric magnitude of the brightest giant in each cluster with theoretical calculations of the luminosity of the helium core flash, finding acceptable agreement. The agreement is significantly worse, however, if the globular cluster distance scale advocated by Sandage is used instead. Further, we point out that the "semiempirical" method used to generate the giant branch data included with the Revised Yale Isochrones is valid only for a particular choice of distance scale. On this scale, which is not that of the horizontal branch models, the observed and theoretical giant branches agree reasonably well, though the deviations in color at high luminosities remain substantial.
