GaugeCam Remote Image Manager Educational 2 (GRIME2) is a free, open‐source Windows application for water level measurement in surface water bodies using ground‐based time‐lapse imagery. We evaluated a new “stop sign” (i.e., octagon shape) calibration target and GRIME2 algorithms by comparing results to other measurement methods over a 26‐month period. GRIME2 measurements were closely matched to visual observations of a staff gauge in images (Staff) and a traditional water‐level sensor (Transducer). GRIME2 and Transducer measurements were within ±0.1 ft (3.0 cm) of Staff measurements for 95% and 90% of paired measurements, respectively. After removing time periods when biofouling, ice and/or snow were visible on the GRIME2 target background, 98% of GRIME2 and Staff measurements were within 3 cm. Error metrics for all paired GRIME2 and Transducer measurements (70,668 pairs) included root mean square error (RMSE), absolute error (MAE), and Nash Sutcliffe Efficiency (NSE) of 3.6 cm, 2.3 cm and 0.83, respectively. With biofouling, ice and/or snow periods removed, error metrics improved to 2.1 cm, 1.5 cm, and 0.95, respectively, for 52,570 paired measurements, with 90% of paired measurements within 3 cm. The new GRIME2 greatly eases field setup and image processing. Only octagon facet length and a reference water level are needed to measure water level. Calibration is done on every image to account for camera movement, and the command line interface supports batch and automated processing for data dashboards. With proper calibration target placement and maintenance, GRIME2 can produce robust water level measurements over relatively long observation periods.