Opioids are commonly prescribed to manage acute postoperative pain. However, individuals vary in their opioid dosing needs, with no validated biomarkers to guide prescribing. We used electronic health records (EHR) data from the Million Veteran Program sample to investigate individual differences in opioid analgesic dosing following knee (n=18,540) or hip (n=9,363) arthroplasty. We extracted data from pharmacy records on inpatient opioid medications to estimate average daily morphine milligram equivalent (MME) dosages and used genome-wide association studies (GWAS) to identify genetic variants associated with daily MME dosage, followed by a cross-ancestry and cross-phenotype meta-analysis. We identified two genome-wide significant loci in the African ancestry knee arthroplasty GWAS and three in the Admixed-American ancestry hip arthroplasty GWAS. Multiple other loci in African, Admixed-American, and European populations were nominally significant. The study provides a framework for the use of EHR data to examine the genetics of opioid dosing in post-operative care.