Small wind turbines are often sited in more complex environments than the open terrain sites assumed in relevant installation guidelines or in the international small wind turbine design standard IEC61400-2. The built environment is an example of such a complex environment and installation of small wind turbines on the rooftops of high buildings has been suggested by architects and project developers as a potential means of incorporating sustainable energy generation into building design. In the absence of guidelines for installing wind turbines in the built environment, two key wind measurement parameters are the rate at which a data acquisition system (DAQ) samples the sensor, and the period over which the sampled data is averaged. This paper presents the results of the effect of sampling rate and averaging period on turbulence measurements from a monitoring system on a building rooftop, in order to inform the process of developing guidelines. The results will inform the development of a Recommended Practice of wind resource assessment in the built environment, via the International Energy Agency Task 27. The key finding of the paper is that, in general, 10 Hz sampling and 10 min averaging period give upper estimates for turbulence intensity and maximum values of the turbulence power spectra. Using these conservative values in the design of the turbine may be the best approach to ensure that the turbine can handle both the fatigue loads and resonance due to gusts. (C) 2014 Elsevier Ltd. All rights reserved.