The wind turbine operational characteristics, power measurements and meteorological measurements from Horns Rev offshore wind farm have been identified, synchronized, quality screened and stored in a common database as 10 min statistical data. A number of flow cases have been identified to describe the flow inside the wind farm, and the power deficits along rows of wind turbines have been determined for different inflow directions and wind speed intervals. A method to classify the atmospheric stability based on the Bulk-Ri number has been implemented. Long-term stability conditions have been established, which confirms, in line with previous results, that conditions tend towards near neutral as wind speeds increase but that both stable and unstable conditions are present at wind speeds up to 15 m s −1. Moreover, there is a strong stability directional dependence with southerly winds having fewer unstable conditions, whereas northerly winds have fewer observations in the stable classes. Stable conditions also tend to be associated with lower levels of turbulence intensity, and this relationship persists as wind speeds increase. Power deficit is a function of ambient turbulence intensity. The level of power deficit is strongly dependent on the wind turbine spacing; as turbulence intensity increases, the power deficit decreases. The power deficit is determined for four different wind turbine spacing distances and for stability classified as very stable, stable and others (near neutral to very unstable). The more stable the conditions are, the larger the power deficit. Copyright © 2011 John Wiley & Sons, Ltd.