This study examines office return determinants for major European office centers during the period 1990-2006. Our main hypothesis is that while the demand for office space is regarded as derived demand, the state of the local economy explains more about local office markets than the state of the aggregate national economy. By means of a dynamic panel data model we test whether the local nature of office markets makes a model based on national data biased if the determinants of office returns do not move in the same direction for each of the individual local office markets when compared to the national aggregate. We test this pivotal shift/share problem in a distinct way by employing a unique dataset which includes disaggregated, or city level, and national variables. This dataset allows us to include local data, supplemented with national data for variables which are purely national in nature to study local office markets and makes a comparison with previous studies based on national level data possible. The data used in this study include real estate data by Jones Lang LaSalle and European economic variables based on Eurostat data. Besides panel data analysis on the total sample of cities we also test for sub-samples based on office market size and economic circumstances. Our results provide insight into the applicability of aggregate data for office market research versus the application of more cumbersome and less frequently available local data.