Ground clutter (GC) is received when the mainlobe or sidelobes of the antenna illuminate objects on the ground. Weather radar data measured at low elevations can be highly contaminated by ground clutter. It is desirable to identify clutter for high quality of weather data. A method of scan-to-scan correlation is proposed and studied. Weather signals are correlated from pulse to pulse but not from scan to scan because the common correlation time of weather signals is around 10 ms while the time for NEXRAD to complete an azimuthal scan at a low elevation is around 32 s. However, ground clutter signals from the same location but observed on two adjacent scans can be highly correlated. Thus, the correlation time of GC and weather are significantly different. Furthermore, the dual-scan data is available from NEXRAD radars for observations at the lowest elevation angles. For example, in the VCP 21 mode, the radar scans twice at elevation angles at 0.5° and 1.5° with different PRTs (Pulse Repetition Time). Cross-correlation coefficient between these dual-scan time-series data collected at the same elevation is calculated and used to distinguish clutter from weather signals. The method is demonstrated with data collected by KOUN radar – a prototype for NEXRAD with dual-polarization capability, and by the NWRT (National Weather Radar Testbed) phased array radar.