The spatial structures of turbulent flow in the atmospheric boundary layer (ABL) are complex and diverse. Multi-point spatial correlation measurements can be used to improve the understanding of these structures and their statistics. In this context, we investigate Taylor’s hypothesis and the statistics of spatial structures on the microscale in this study. For the first time, simultaneous horizontally distributed wind measurements with a fleet of 20 quadrotor UAS (unmanned aerial systems) are enabled. The measurements were taken at different heights and under different atmospheric conditions at the boundary layer field site in Falkenberg of the German National Meteorological Service (DWD). An adaptable horizontal flight pattern has been especially developed, consisting of measurements distributed along and lateral to the mean flow direction with separation distances of 5...205 m. The validity of Taylor's hypothesis is studied by examining cross-correlations of longitudinally distributed UAS and comparing them with autocorrelations of single UAS. To assess the similarity of flow structures on different scales, the lateral and longitudinal coherence of the streamwise velocity component is examined. Two modeling approaches for the decay of coherence are compared. The experimental results are in good agreement with the model approaches for neutral atmospheric conditions, whereas in stable and convective ABL, the exponential approaches are not unconditionally valid. The validation results and the agreement with the literature on coherence in the ABL underline the potential of the UAS fleet for the purpose of spatial turbulence measurements.