A blind quality control (BQC) program in blood alcohol analysis was implemented at the Houston Forensic Science Center (HFSC) in September 2015. By mimicking authentic toxicology blood evidence, the laboratory can perform a concurrent evaluation of their technical and administrative casework procedures and test the accuracy and reliability of their volatile analysis method in a format that is blinded to the analyst. From September 2015 to November 2023, HFSC's Quality Division submitted 1228 antemortem whole blood samples: 292 ethanol-negative samples and 936 ethanol-positive samples at sixteen target concentrations (0.051, 0.080, 0.100, 0.110, 0.120, 0.130, 0.150, 0.160, 0.170, 0.180, 0.190, 0.200, 0.230, 0.240, 0.250, and 0.260 g/dL). A second, unopened blood tube in 168 of the 1228 BQCs was also analyzed after 721-1140 days: 24 ethanol-negative samples and 144 ethanol-positive samples at five target concentrations (0.080, 0.100, 0.130, 0.180, and 0.240 g/dL). All 316 ethanol-negative samples remained negative. After 42-758 days, the average (median, range) change in ethanol concentration of the 936 positive samples was -1.4% (-1.3%, -12.0% to +8.4%) with a statistically significant difference (P < 0.001) observed for the gradual decline in blood alcohol concentration (BAC) over time. The average BAC percentage differences per target concentration, ranged from -6.4% (-0.008 g/dL) to +5.7% (+0.011 g/dL), were within HFSC's current measurement uncertainty (9.4% at k=3), showing no apparent correlation between the change in ethanol and the theoretical target concentration. As the analysis time between the two blood specimens from the same evidence kit extended, the loss in ethanol significantly increased (P < 0.001).