This paper investigates the problems in integrating different data sets for social science research. A cross-national analysis of the effects of family migration on labor market participation rates for female partners was used as an example to concretize these problems. The data used in the cross national analysis came from the 1991 British Census Sample of Anonymised Records and the 1990 United States Public Use Microdata Sample. The problems addressed were the following: data collection; manipulation and reliability; question type and definition; and measurement of individual and family variables. As part of the analysis, some empirical findings were also presented. Significant overall differences between the US and Great Britain (GB) samples were observed, particularly concerning graduate qualifications, marital and employment status, and ethnic group. There were higher numbers of married partners and full-time workers in the US sample, while female partners undertaking part-time employment were greater in GB. Furthermore, the US sample had more couples with a female graduate as compared with the GB sample who had a higher proportion of nongraduate couples. Finally, this paper recognized the importance of cross- national research in the evaluation of national ideological and institutional structures.