ABSTRACT: Epidemiological studies of sperm quality are hampered by problems such as low participation rates and poor comparability of results due to methodological differences in semen analysis. More objective sperm quality‐related serum markers would facilitate worldwide comparisons of male reproductive status. Our objectives were to investigate to what extent a set of hormonal indices of testicular function, previously established in a clinical setting, could predict regional variations in seminal parameters in men from the general population. We recruited 101 men aged 20–40 years from two regions in Flanders, and assessed their sperm parameters and levels of serum reproductive hormones. In one region compared to another, the participants had lower sperm counts (by 34%; P = .06), lower total sperm counts (by 41%; P = .02) and poorer sperm morphology (by 32%; P < .001), which were paralleled by significantly lower levels of free testosterone (by 11%; P = .03), while for total testosterone (T) and follicle‐stimulating hormone (FSH), the differences were not significant (both P = .09) at 10% and 17%, respectively. There were no differences in inhibin B and the T to luteinizing hormone (LH) ratio, which are markers of testicular function. Receiver operating characteristic curve analysis revealed that T:LH, inhibin B, and the inhibin B/FSH ratio had significant discriminatory power between men with sperm concentrations below or above 13.5 × 10 6 /mL. Regional variations in the semen quality of community‐dwelling individuals are not necessarily reflected in altered hormonal indices of testicular function and thus, these markers, validated in clinical settings, are not valid substitutes for the traditional semen quality assessment used in epidemiological population studies.