We investigated the disturbance effects of offshore windfarm construction on harbour porpoises Phocoena phocoena using acoustic porpoise monitoring data and noise measurements during construction of the first 7 large-scale offshore wind farms in the German Bight between 2010 and 2013. At 6 wind farms, active noise mitigation systems (NMS) were applied during most piling events, and 1 was constructed without. Based on generalized additive modelling analyses, we describe a clear gradient in the decline of porpoise detections after piling, depending on noise level and distance to piling. Declines were found at sound levels exceeding 143 dB re 1 µPa2s (the sound exposure level exceeded during 5% of piling time, SEL05) and up to 17 km from piling. When only considering piling events with NMS, the maximum effect distance was 14 km. Compared to 24-48 h before piling, porpoise detections declined more strongly during unmitigated piling events at all distances: at 10-15 km declines were around 50% during piling without NMS, but only 17% when NMS were applied. Within the vicinity (up to about 2 km) of the construction site, porpoise detections declined several hours before the start of piling and were reduced for about 1-2 d after piling, while at the maximum effect distance, avoidance was only found during the hours of piling. The application of first generation NMS thus reduced the effect range of pile driving and led to a lower decline of porpoise detections over all distances. However, NMS were still under development and did not always work with equal efficiency. As NMS have further developed since, future investigations are expected to show additional reduction of disturbance effects.