In the past decade, several parts of central United States, including Oklahoma, have experienced unprecedented seismicity rates, following an increase in the volumes of wastewater fluids that are being disposed underground. In this article, we present a semi‐empirical model to hindcast the observed seismicity given the injection time history. Our proposed recurrence model is a modified version of the Gutenberg–Richter relation, building upon the seismogenic index model, which predicts a linear relationship between the number of induced events and the injected volume. Our methodology accounts for the effects of spatiotemporal pore‐pressure diffusion, the stressing‐rate dependency of the time lag between injection and seismicity rate changes, and the rapid cessation of seismicity upon unloading. We also introduced a novel multiscale regression, which enabled us to produce grid‐independent results of increased spatial resolution. Although the model is generic to be applicable in any region and has essentially only two free parameters for spatial calibration, it matches the earthquake time history of Oklahoma well across various scales, for both increasing and decreasing injection rates. In the companion paper (Grigoratos, Rathje, et al., 2020), we employ the model to distinguish the disposal‐induced seismicity from the expected tectonic seismicity and test its forecasting potential.