Abstract Introduction When the core lab at a large tertiary care hospital introduced an automated laboratory system from a major manufacturer, the automated line aliquoted patient serum into open cups for all chemistry testing. It was subsequently discovered that if sample testing was delayed, electrolyte abnormalities could be introduced from evaporation, resulting in an increased anion gap (presumably due to a loss of bicarbonate with evaporation of carbon dioxide). To address this issue, the core lab changed the sample processing protocol in April 2023 from an open cup system to testing primary tubes on the analyzers. The purpose of this study was to evaluate the impact of switching to a primary tube testing system on artefactually elevated anion gap results. Methods Equivalent ten-week blocks of anion gap, creatinine and sample turnaround time (from receipt on the line to analyzer perform time) results from pre- to post-primary tube conversion were collected from the laboratory information system. Data from weekend testing and from repeat samples from the same patient were excluded. Data was cleaned and analyzed using R. Results Multivariable linear regression showed that when adjusted for the turnaround time and creatinine result, the open cup system was significantly associated (p<0.0001) with a 0.95 [95% CI 0.69-1.21] increase in anion gap. There was a positive relationship between turnaround time and anion gap, with a slope of 0.017 (i.e., a 0.017 increase in anion gap per minute delay; p<0.0001). Overall, the proportion of abnormal results (defined as anion gap >=16) decreased from 26% to 10% after switching to a primary tube testing system. Furthermore, there was an association between the times of day in which the lines were overloaded with specimens, turnaround time, and an increase in abnormal anion gap results. Overall, the mean anion gap decreased by 1 with a turnaround time of one hour and 2 after two hours by switching to the primary tube system. This trend in anion gap was largely driven by a corresponding decrease in bicarbonate. Sodium and chloride levels did not show any relationship with turnaround time. Conclusions Using an open cup system for electrolyte testing can cause spuriously elevated anion gap levels, especially with testing delays. Switching to a primary tube testing system largely resolved the issue; however, the data still showed a weakly positive relationship between turnaround time and anion gap for the primary tube testing system, so additional work is needed to identify potential causes of this relationship.