Earthquakes associated with fluid injection in various geo-energy settings, such as shale gas and deep geothermal energy, have shelved many projects with great potential. However, the injection-rate dependence of earthquake nucleation length, i.e., the slowly slipping (creeping) fault length in preparation for a subsequent earthquake (Kaneko & Lapusta, 2008), remains elusive. In this study, we take a step towards this issue by performing fluid injection experiments on low-permeability granite samples containing a critically stressed sawcut fault at different local injection rates (0.2 mL/min and 0.8 mL/min) and confining pressures (31 MPa and 61 MPa) (c. f., Ji & Wu, 2017; Wang et al., 2020). An array of local strain gauges and acoustic emission (AE) hypocenter locations were used to monitor the precursory slip of critically stressed faults before injection-induced stick-slip failure (c. f., Passelègue et al., 2020; Wang et al., 2020). The nucleation length was determined for each injection-induced stick-slip event, and its dependence on effective normal stress and injection rate was explored. Herein, we compile the processed data obtained from the experiments in four Excel worksheets. The full description of the methods is provided in Ji et al. (2022).