Historical demographers have extensively studied the importance of nuptiality regulating marital fertility and population growth in pre-industrial Europe. Most of the focus has been on the timing and incidence of the marriage but also on the differences of partner selection. This latter strand of research mainly engages with social boundaries in marriage formation, age and geographic homogamy. However, most of the previous research on marriage patterns and family formation has largely focused on Western and Northern Europe, meaning that a considerable part of the continent has not been included in these discussions. In some of the less studied regions in Eastern Europe, the existing knowledge on demographic modernisation, including changes in nuptiality, stems exclusively from aggregate-level analyses.

This paper aims to contribute to the literature by analysing heterogamous versus homogamous partner selection in the late 19^th century urban marriage market in Tartu, Estonia, using the First Russian Imperial census in 1897. We ask whether educationally heterogamous marriage was more associated with certain groups, potentially contributing to their social mobility. The analysis focuses on four main characteristics: socio-economic background (education), ethnicity, age, and place of birth. We apply different regression models to estimate the probability of ending up in a heterogamous union and to analyse the age gap between spouses. With churches' marriage records linked to the census data, transition to first marriage can be analysed to estimate the determinants of the timing and probability of homogamous and heterogamous marriages. Preliminary results indicate a higher proportion of heterogamous unions among individuals with medium or high education. The results also suggest that a higher than primary education and foreign background made it more likely that men were able to marry younger wives.