This study provides a detailed analysis to regional office real estate markets in the United Kingdom. A vector error correction (VEC) approach is applied to a unique panel dataset that covers the time period from 1981 until 2004 and allows a disaggregation to the NUTS 2 level. Long-run equilibrium relationships and shortterm corrections among total returns and the key macroeconomic variables gross domestic product (GDP), total investment and unemployment are captured. Different samples are used to control for the special role of the London market and to alleviate potential bias due to differences in the number of properties within regions. The results leave little ground for assuming that the economic variables have no impact on total office returns. They rather provide evidence that there are relatively strong long-run relationships. The assumption is supported that the longrun relationships are causal and running from the economic variables to total return. Furthermore, there is evidence for short-term causal relationships between economic variables, in particular total investment, and total return, as well as for total returns adjusting to the long-term disparities resulting from changes in the variables. Consequently, the economic variables do not only seem to provide short-term information but also short-term immediate effects on the movements of total returns. The character of the long-term equilibrium and short-term corrections, however, is not identical across samples while London indeed seems special.