These data are from Laboratory Earthquake Experiments from the Cornell 3 m apparatus in support of the following research: Loading a 3-meter granite slab containing a saw-cut simulated fault, we generated rupture events that spontaneously nucleate, propagate, and arrest before reaching the ends of the sample. These rupture events have a slip distribution that varies along the fault and make them more similar to natural earthquakes than standard stick-slip events that rupture the entire sample. Through LEFM (Linear Elastic Fracture Mechanics), we showed how the balance between energy release rate and fracture energy governs the termination of a rupture. In our experiments, fracture energy is essentially constant compared to the orders-of-magnitude variations in energy release rate so ruptures terminate because they run out of available strain energy. The utility of the model for both 3-m rock experiments and 200-mm PMMA experiments, and the similarity of fracture energy coefficient between the two materials, verifies the adequacy of PMMA as an analog to crustal rock in this context. Finally, the LEFM-based model provides a framework for linking friction properties and on-fault stress conditions to observable earthquake sequences