This paper concerns experimental investigation of two-phase air-water flow in a 5 mm circular channel for different gas and water volume flow rates. Flow patterns are recorded using a high-speed digital camera. To investigate the spatial correlations of flow structures, we propose a new technique for extracting high resolution space-time series from video frames. For the extracted series, Eulerian space-time correlations are estimated and represented by correlation curves and spatio-temporal 2D maps for several gas and water volume flow rates as well as for different channel inclinations. The characteristic features of the flow patterns can be deduced from the correlation results. (C) 2016 Elsevier Ltd. All rights reserved.