<p>Assessing the temperament of horses is important to ensure their welfare, management, and safety. This study evaluated the plasma-saliva relationships for cortisol, oxytocin, and serotonin, and explored their associations with temperament traits in 40 horses using enzyme-linked immunosorbent assays. Significant correlations were observed between cortisol (r² = 0.441) and oxytocin (r² = 0.587) concentrations, whereas serotonin showed a weak correlation (r² = 0.133). Bland-Altman analysis revealed acceptable agreement for cortisol but substantial individual variability for oxytocin, despite its stronger correlation, and concentration-dependent bias for serotonin. Association between hormone and temperament were examined exploratorily in the Jeonju cohort only (<i>n</i> = 20) using Spearman’s rank correlations with Benjamini-Hochberg false discovery rate correction. Several nominal associations were observed at the unadjusted level, but none remained significant after correction for multiple testing. Horses with regular human handling showed lower cortisol and higher oxytocin concentrations than minimally handled horses; however, these between-group differences should be interpreted cautiously because they were confounded by management context. Overall, salivary cortisol showed the most practical potential as a non-invasive indicator in horses. In contrast, oxytocin and serotonin findings should be interpreted cautiously, and further analytical validation and larger, adequately powered studies are needed before their individual-level application to temperament assessment can be supported.</p>