This work analyses the weak light performance efficiency of six different free field and grid-connected PVarrays of various module technologies (mc-Si, CIGS, CdTe). The installed capacities per array range from 1,8kWp to 2,2kWp. The studied systems are located at the same site in Germany and are mounted on a ventilated and shaded-free structure. The irradiance was measured with both a pyranometer and a c-Si reference cell oriented at the plane of the array. Besides the weather conditions, the arrays share a similar system configuration (same inverter model, rack type and orientation, temperature sensors, etc.) making their results highly comparable. The data collected comprised 312 days from September 2011 until July 2012. When analyzing the data points at moments of low irradiance (< 800W/m2), a high dispersion was found which is most likely due to optical effects, different spectral distributions and changing module temperatures. To reduce the optical and spectral effects, a series of data filters were applied to limit the points used (Air Mass ≤4 and Angle of Incidence ≤50°). To compensate for the temperature effects and translate the values to STC (25°C), five different methods were assessed. The Procedure 2 of the IEC 60891 was considered the most suitable due to its relative simplicity, availability of parameters in the datasheets, good accuracy even with missing values, and the potential to improve the results when the complete set of inputs is available. The final analysis of the weak light performance showed no clear superiority of any particular PV-technology. From the results of this study, an overall advantage of thin-film over crystalline modules cannot be anymore concluded.