In 2010, the American Community Survey (ACS) collected data simultaneously with the decennial census. Comparisons of the number of vacant and occupied housing units in the 2010 Census and the 2010 ACS 1-year estimates for the U.S. show statistically significant differences. Other statistics derived from these measures are statistically different at the national and state levels, too (Griffin, 2011). For example, the national ACS vacancy rate of 13.1 percent, ± 0.1 percentage points, is greater than the 2010 Census vacancy rate of 11.4 percent. While it may seem contradictory that the 2010 Census and ACS results would vary, there are several reasons why this may be (Cresce, 2011; Griffin, 2011). In a study that compared the final statuses of addresses in the full 2010 ACS sample with their outcomes in the 2010 Census, Anderson and Hefter (2011) found differences in the Census and ACS frames in addition to inconsistency in occupancy statuses for addresses that both programs had recorded as housing units. The inconsistently classified housing units were marked occupied in the Census and vacant in the ACS twice as often as those marked vacant in the Census and occupied in the ACS. This is one of the main reasons for a higher ACS vacancy rate. This report supplements those findings by studying inconsistent classifications by geography, population and housing characteristics, and operational factors associated with the 2010 Census and ACS data collections. To improve the interpretation of the results, we designed this analysis so that the time differences between the dates of Census enumeration and ACS interview are limited.