As the most destructive seismic episode ever known in eastern Taiwan, the 1951ML7.3 Hualien – Taitung earthquake series consisted of sequential ruptures along four distinct fault segments. It provides a good opportunity to study earthquake triggering processes along an active fault at an oblique arc‐continent collision boundary. This sequence initiated on 21October 1951 with theML7.3 Hualien main shock and a group ofM6+ aftershocks nearby. TheML6.0 Chihshang earthquake occurred 34 days later and 100 km away from the main shock. TheML7.3 Yuli earthquake followed 3 m later and 5 km away from the Chihshang event. Two days later, theML6.0 Taitung earthquake shocked a region 40 km away from the precedingM6 event and completed the sequence. The first triggered rupture outside the main shock area did not occur on the nearby Yuli fault segment but occurred 100 km away at the Chihshang fault. Calculations of static Coulomb stress change show that most of the major aftershocks were located in areas of enhanced static stress change. However, the stress transfer alone cannot explain triggering across 100 km. With the rate/state stress transfer model, we computed the temporal order of encouraged ruptures on different segments along the collision boundary. The results show that 34 days following the major shocks in Hualien, the Chihshang segment had a higherM6+ (M≧ 6) earthquake probability due to its significantly higher (at least an order of magnitude) background seismicity rate than the other two segments. After the Chihshang event, the rate/state model predicted a higherM6+ earthquake probability in the Yuli segment, also matching the observation. In this case, the Yuli segment was triggered ahead of the Taitung segment because of its larger increase in Coulomb stress change.