The myeloperoxidase index (MPXI), on ADVIA hematology analyzers, reflects the mean neutrophil myeloperoxidase staining. It is used as a marker of inflammation in animals and people, but assay variability and storage stability are unknown. We aimed to determine MPXI precision and stability with refrigerated storage of canine and equine EDTA-anticoagulated blood and compared MPXI results between two analyzers. Inter-assay coefficients of variations (CVs) were determined from three human-based controls assayed before and after a 20- or 21-day calibration. Blood from 14-16 dogs and 26 horses was assayed 4-10 times within 1day for intra-assay CV measurements. Median control and single run results from 18 canine and 35 equine samples were compared between analyzers. Blood from 10-12 dogs and 10-11 horses was analyzed after collection, and 24, 48, and 72hours of refrigerated storage. Inter-assay CVs of controls were 10.7%-15.9% and 6.4%-9.6% before and 4.3%-7.7% and 2.8%-17.5% after calibration, for ADVIA 1 and 2, respectively. Calibration altered peroxidase gain settings and improved precision. Intra-assay CVs were 0.6%-64% and 3%-350% for canine and equine samples, respectively. Median MPXI results differed significantly between the analyzers, likely from calibration-associated changes in gains. MPXI decreased with storage, and with variable changes between animals and analyzers. Platelet clumps and lipid contributed to the variability in replicate MPXI measurements. MPXI has a higher variability in equine samples than in canine samples. Equivalent results might not be obtained between analyzers. Results change unpredictably with repeated analyses over 72hours. MPXI measurements might only be useful in controlled research settings.