The long-term downward trend in the percentage of extended family households in the U.S. came to a halt during the 1980s, a change that coincided with a growing gap between immigrants and natives in the percentages of households adopting extended family structures. Using 1970, 1980, and 1990 census data, this research assesses the degree to which changes in the volume and composition of immigration have contributed both to the increase in the proportion of the U.S. population residing in extended family households and to the widening gap between immigrants and natives. Our results demonstrate that immigration explains only a little of the total increase in extended living arrangements in the total population, but that the increasing differential between immigrants and natives during the 1980s resulted from increases in horizontally extended households among immigrants. Mexican, Guatemalan, and Salvadoran immigrants accounted for most of this increase, primarily because of increases in the proportion of young, single adults living with relatives and increases in poverty rates among immigrants from these countries. Key Words: extended family, household composition, immigration. One important focus of postwar American social science has involved studying changes in family household structures and their determinants, including changes in extended family living arrangements or the tendency for relatives to coreside in the same household (Cherlin, 1992; McLanahan & Caspar, 1995; Ruggles, 1994). Studies of changes in family household structures in the U.S. generally have documented a longterm downward trend during the 20th century in the prevalence of extended family households (Goldscheider & Goldscheider, 1989; Sweet & Bumpass, 1987). For example, among Whites in the U.S., households containing nonnuclear relatives dropped continually from 20% in 1910 to a low of 7% in 1980; among Blacks, the percentages have fluctuated from 24% in 1910 to 17% in 1980 (Ruggles, 1994). During the 1980s, however, the decline in extended living arrangements came to an end. Among all households, those containing extended family members increased from 10% in 1980 to 12% 1990, a shift that raises the question of whether a fundamental change has occurred in the factors affecting the commonality of extended family living arrangements. What might explain recent changes in the prevalence of extended family households? One factor is immigration. During the 1970s and 1980s, the U.S. experienced dramatic shifts in the volume and composition of immigrants (Edmonston & Passel, 1994). Average annual immigration increased from 332,000 people per year during the 1960s to 449,000 people per year during the 1970s and 734.000 people per year during the 1980s. Immigrants coming from Canada and Europe declined from 68% of all arrivals during the 1950s to 13% during the 1980s, whereas those from Asia and Latin America increased from 31 % to 84% over the same period (U.S. Immigration and Naturalization Service, 1995). Moreover, among immigrants, extended family households have increased steadily over the past 30 years, growing from 16% of recent immigrants coresiding as dependents in other relatives' households in 1960 to 30% in 1990 (Biddlecom, 1994) and thereby increasing the gap in extended family living arrangements between immigrants and natives. The rise in the number and proportion of immigrants from less developed countries where extended family structures are more common (De Vos, 1995) thus constitutes a potential explanation of (a) the recent increase in the percentage of extended family households in the total U.S. population, (b) the 30-year increase in the percentage of extended family households among immigrants and the growing gap in the percentage of family households between immigrants and natives, which we term the nativity gap. Insight into the extent to which immigration has affected coresidential relationships also can be obtained by examining different types of extended family household structures. …