The effects of population size and growth rate on population forecast accuracy have been well documented. For example, we know that small places generally have larger errors than large places; that errors are generally higher for places with high growth rates than places with low growth rates; and that size of error generally remains more stable over time than does the direction of error. In this paper, we delve more deeply into these relationships using data for 2,482 counties in the United States and expand the analysis to include a third explanatory variable, prior forecast error.