Some useful approximations are developed in this paper that describe large-scale epidemic phenomena for the deterministic and stochastic Reed-Frost model. These characteristics provide an a priori quantitative description of the epidemic curve for the deterministic case, such as a threshold requirement, the total size of the epidemic, and the degree of skewness in the epidemic curve. A Poisson distribution asymptotically describes the distribution of the total size of an epidemic when the relative removal rate is well below threshold for the Reed-Frost chain binomial model. These properties are established by an extension of the large-scale epidemic phenomena of the Kermack-McKendrick model.