Objective: To examine whether the association between women's origin and early pregnancy overweight and obesity (OW/OB) varies by length of residence in Sweden.

Design: This cross-sectional observational study used pooled Swedish population register data from 1992 to 2012. Logistic regression models were run to estimate odds ratios (OR) of early pregnancy OW/OB and 95 % confidence intervals (95 % CI), comparing Swedish-born and immigrants by length-of-residence categories while adjusting for covariates.

Setting: Sweden.

Participants: In total, 1 771 821 pregnant women, 315 992 of whom were immigrants.

Results: With longer residence in Sweden, more immigrant women from various origins exhibited higher odds of experiencing early OW/OB compared with Swedish-born women. Findings specifically showed increased odds of early pregnancy OW/OB with increasing length of residence for women born in Latin America, Europe-27 and Southeast Asia. For example, immigrant women from Latin America residing in Sweden for < 6 years showed similar odds as Swedish-born (OR <= 5 years 0 center dot 92, 95 % CI 0 center dot 87, 0 center dot 98), while their longer residing counterparts showed higher odds than Swedish-born women (OR(6-15 years)1 center dot 21, 95 % CI 1 center dot 14, 1 center dot 28 and OR >= 16 years 1 center dot 68, 95 % CI 1 center dot 59, 1 center dot 78). Mixed results were found for other origins.

Conclusions: The current study suggests that host country conditions might play an important role in explaining OW/OB among some groups of immigrant women. Although further studies are needed to disentangle the mechanisms that generate these health inequalities, policy efforts should focus on immigrant reception and early integration to prevent pregnancy-related OW/OB.