Measurements of tibial acceleration during running must be reliable to ensure valid results and reduce errors. The purpose of this study was to determine the reliability and minimal detectable difference (MDD) of peak axial and peak resultant tibial acceleration during overground and treadmill running. The authors also compared reliability and MDDs when peak tibial accelerations were determined by averaging 5 or 10 trials. Tibial acceleration was measured during overground and treadmill running of 19 participants using a lightweight accelerometer mounted to the tibia. Peak axial and peak resultant tibial accelerations were determined for each trial. Intraclass correlation coefficients determined within-session reliability, and MDDs were also calculated. Within-session reliability was excellent for all conditions (intraclass correlation coefficients  = .95–.99). The MDDs ranged from 0.6 to 1.4 g for peak axial acceleration and from 1.6 to 2.0 g for peak resultant acceleration and were lowest for peak axial tibial acceleration during overground running. Averaging 10 trials did not improve reliability compared to averaging 5 trials but did result in small reductions in MDDs. For peak axial tibial acceleration only, lower MDDs indicate that overground running may be the better option for detecting small differences.