Downsized direct-injected boosted gasoline engines with high specific power and torque output are leading the way to reduce fuel consumption in passenger car vehicles while maintaining the same performance when compared to applications with larger naturally aspirated engines. These downsized engines reach brake mean effective pressure levels which are in excess of 20 bar. When targeting high output levels at low engine speeds, undesired combustion events called pre-ignition can occur. These pre-ignition events are typically accompanied by very high cylinder peak pressures which can lead to severe damage if the engine is not designed to withstand these high cylinder pressures. Although these pre-ignition events have been reported by numerous other authors, it seems that their occurrence is rather erratic which makes it difficult to investigate or reliably exclude them. This paper describes a systematic engine dyno testing approach to force the engine into pre-ignition in order to study and characterize these events. A sensitivity study of various parameters shows that pre-ignition can occur repeatedly at the same load levels if boundary conditions are controlled sufficiently, meaning pre-ignition occurrence is less erratic than previously thought. Several hundred pre-ignition events have been recorded and analyzed in this study. A post-processing tool was developed and applied to analyze and characterize all recorded pre-ignition events. The knowledge gained out of these investigations will help to better understand the pre-ignition phenomena and what combustion development activities need to be applied in order to avoid or counteract pre-ignition during an engine development program or afterwards during customer usage in a passenger car.