While a great many state-level longitudinal measures of public opinion are now available, few measures exist at the Metropolitan Statistical Area (MSA) level. I demonstrate a method for estimating public opinion annually in the fifty-three largest MSAs via disaggregation from the Cooperative Congressional Election Study (CCES). Using this method, I construct an MSA-level citizen political ideology measure from 2006 through 2016. This measure is strongly correlated with MSA-level results from the Presidential Elections in 2008, 2012, and 2016. While the MSA ideology scale itself has many potential applications for future research, so, too, does the underlying method of estimating MSA-level measures using CCES data.