Background: A consensus on an accurate method to measure gamma-glutamyltransferase (GGT) activity for clinical purposes has not been achieved among practicing clinical laboratories. To improve analytical trueness, we evaluated the influences of matrix effects in proficiency testing (PT) materials on the measurement of GGT activity in human serum samples. Methods: Five fresh frozen human samples (FFS1-5) and five lyophilized proficiency testing materials (Lyo1-5) were distributed to 23 participating clinical laboratories for the measurement of GGT activity. Target GGT activity values for the samples were obtained by using previously approved reference methods. The results obtained by the Beckman Coulter Unicel DxC 800 Synchron analyzer were compared to the target values assigned by two reference laboratories, and the commutability of the lyophilized materials was evaluated according to Clinical and Laboratory Standards Institute (CLSI) guideline EP14-A2. Results: The relative bias between the results obtained by the Beckman Coulter analyzer and the reference target values ranged from -27.2% to -18.0% for FFS1-5 and from 9.1% to 2.5% for Lyo1-5. Non-commutability of all lyophilized samples falling outside of the 95% prediction interval was observed. Conclusions: The results obtained for the lyophilized PT materials were deemed acceptable within the total allowable errors, suggesting that matrix effects may impart a false sense of confidence that clinical analytical systems are performing very well. A primary reference measurement procedure on fresh frozen serum provides a valuable method for evaluating the trueness of results measured by PT.