Observations made during forensic automotive crash investigations have identified instances of non-zero, post-crash speedometer readings and created questions as to the validity of the indicated speed relative to the vehicle speed at impact. Previously published work has addressed many issues related to the reliability of non-zero, post-crash speedometer readings identified in vehicles as well as motorcycles. Much of this work established criteria that related the reliability of the post-crash needle position to the design of the stepper motor that controls the needle. Part of this criteria is related to the static torque associated with the speedometer needle shaft rotation due to outside (crash) forces. The published criteria were evaluated in staged crash tests which investigated the ability to maintain needle position under longitudinal and lateral forces after an electrical power loss. In an effort to extend the science, this paper compares non-zero, post-crash speedometer readings with event data recorder (EDR) data from twenty-one real-world crashes where both non-zero, post-crash speedometer readings and EDR data were available. Results from this study suggest a positive correlation between non-zero speedometer readings and vehicles experiencing both an electrical power loss and a single impact. However, this study also shows a negative correlation between non-zero speedometer readings and vehicles experiencing an electrical power loss and multiple impacts. Eighteen of the twenty-one vehicles had valid pre-crash EDR speeds. Of those eighteen vehicles, the speedometer position and EDR speed were within ±15% for fourteen of the vehicles. The current study also demonstrates that non-zero tachometer readings do not always improve confidence levels of non-zero speedometer readings. While the post-crash needle position may provide a good estimate of the travel speed of the vehicle at the time of power loss, there are numerous other factors which must be considered prior to accepting these readings.