BACKGROUND: Women are under-represented in cardiovascular research, leading to poorer outcomes. Investigating sex-differences in electromechanical function is essential for improving therapy evaluation. This study presents sex-specific human cellular and biventricular electromechanical models for mechanistic investigation of sex-differences in therapeutic response. METHODS: Protein genomic expression data from healthy human myocytes calibrated sex-specific electrophysiological models, integrated into biventricular models with male and female anatomies. Multi-scale validation utilised sex-specific clinical and experimental datasets, including responses to drug action. Ionic mechanisms underlying sex-differences in drug response were explored. RESULTS: Simulations showed agreement with clinical ECGs, including QTc intervals (Male: 312 ms; Female: 339 ms), and T-wave amplitude (6-9 % difference). Mechanical biomarkers (LVEF, Female: 68 %; Male: 50 %) matched sex-stratified UK Biobank data (n = 806; 46 % Male). ECG sex-characteristics were driven by ionic differences, while mechanical differences stemmed from anatomical and ionic differences. Simulations predicted exacerbated QTc prolongation under Dofetilide in women (54-78 % higher than males) and T-wave amplitude reduction in men (max: -0.25 mV). Verapamil increased T-wave amplitude in females and decreased it in males, without affecting QTc. Simulations demonstrated reduced repolarisation reserve and increased QTc susceptibility in women via hERG inhibition, while enhanced calcium buffering protected against T-wave amplitude loss. LVEF changes in response to calcium block were more sensitive to anatomical differences between male and female than to ionic sex phenotypes. CONCLUSION: Sex differences in repolarisation reserve, calcium handling, and anatomy are key factors underpinning ECG and LVEF responses to drugs. Specifically, under calcium block, females showed more preserved LVEF, while under hERG block, females showed more QTc prolongation.