Marital isonymy is frequently used to estimate inbreeding and the repeated pairs method is useful to investigate whether the population under examination has subdivisions. These methods can also be applied to registers, such as population censuses, where both spouses' surnames are noted. In this paper, the 1795 census for Córdoba province is analysed. Numerically speaking, Spanish and mixed-race people are the major ethno-social groups in the register. In order to estimate inbreeding, the isonymic method was applied to both groups, at provincial and at parish level. To appreciate to what extent the parishes were genetically isolated, Wright's Fst was also calculated. The repeated pairs method was also used for both groups to assess if population subdivision existed in the units under study. Finally, to evaluate whether the subdivision based on surnames reflected the ethno-social stratification, the same method was used considering the two groups together. At the provincial scale, both groups displayed low inbreeding and micro-differentiation, although the former was higher for the Spanish and the latter for mixed-race groups, which could indicate a more marked conjugal selectivity in the Spanish. At the parish scale, preferences for isonymic spouses were not pronounced either in Spanish or in mixed-race groups; in the Spanish group population subdivision prevailed, with the opposite occurring in the mixed-race group. The estimations from repeated pairs, taking the two groups together, indicated that for the studied populations the surnames do not allow the two groups to be differentiated into isolated reproductive units.