OBJECTIVE: To enhance the quality of secondary data analysis, we compared the usability and the accuracy of free graph digitization software to reproduce published Kaplan Meier (KM) curves. METHODS: A Google search identified 13 different free graph digitization software tools. Of those available on Windows 10, we reviewed several popular tools, including GetData (GD), Graph Grabber (GG), Plot Digitizer (PD) and WebPlotDigitizer (WD) using published KM curves. We evaluated their characteristics for versatility, user-friendliness, and accuracy. Versatility encompassed supported file types and operating systems, and whether the tool could operate as a stand-alone program. User-friendliness included types of graphs that could be detected, flexibility for tracing the graph, and language support. Accuracy was evaluated by comparing median overall survival (OS) and hazard ratios (HR) between the originally published data and the reconstructed summary data after auto-trace digitization. RESULTS: Most software could open any file type. GG and GD worked only with Windows. PD could run on Java 1.6 or later and needed a separate software for auto-trace. GD had trouble detecting colored graphs. PD accepted unlimited undo/redo but the others did not. GG allowed the user to adjust individual points after auto-trace and perform visual confirmation within the same software. GD supports 14 languages while WD supports four languages and others only supported English. GG had the smallest difference for the median OS between the published results (8.60 and 5.30 months) and the reconstructed results (8.60 and 5.39 months). GD had the smallest difference in HR (0.48 vs. 0.49). CONCLUSIONS: Based on our analysis, WD showed the most versatility whereas GG was the most user-friendly. GG and GD had slightly more accurate results than others using auto-trace. Selection of the most appropriate tool may be dependent on the specific desired tool attributes and digitization purpose.