Gross domestic product (GDP) and gross domestic income (GDI) are in theory estimates of the same concept, namely economic production over a defined span of time and space. Yet the two measures are compiled using different source data, and the two measures often give different indications of the direction of the economy. This raises the issue of which of the two measures is a more accurate estimate of economic production. In this paper we present a time-series statistical framework for addressing this issue. Our findings indicate that the latest vintage of GDP has been a better measure of true output over the 1983-2009 period than the latest vintage of GDI. Our model also implies an optimal weighting of GDP and GDI can yield a more accurate estimate of economic output than either GDP or GDI alone. Our empirical findings indicate that a weighting of approximately 60% to GDP yields the best estimate for the 1983-2009 period. When we consider vintages of estimated output, we find that GDI often contains additional information to GDP regarding true output.