Objectives: To determine whether family structure (polygamous or monogamous) is associated with sexual activity among school students in Nigeria. Design: Cross sectional school survey with a two stage, clustered sampling design. Participants: 4218 students aged 12–21 years attending 39 schools in Plateau state, Nigeria. Responses from 2705 students were included in the analysis. Main outcome measure: Report of ever having had sexual intercourse. Variables of interest included sexual history, age, sex, religion, family polygamy, educational level of parents, having a dead parent, and sense of connectedness to parents and school. Results: Overall 909 students (34%) reported ever having had sexual intercourse, and 1119 (41%) reported a polygamous family structure. Sexual activity was more common among students from polygamous families (42% of students) than monogamous families (28%) (χ2=64.23; P<0.0001). Variables independently associated with sexual activity were male sex (adjusted odds ratio 2.52 (95% confidence interval 2.05 to 3.12)), older age (1.62 (1.24 to 2.14)), lower sense of connectedness with parents (1.87 (1.48 to 2.38)), having a dead parent (1.59 (1.27 to 2.00)), family polygamy (1.58 (1.29 to 1.92)), lower sense of connectedness with school (1.25 (1.09 to 1.44)), and lower educational level of parents (1.14 (1.05 to 1.24)). Multistep logistic regression analysis showed that the effect of polygamy on sexual activity was reduced by 27% by whether students were married and 22% by a history of forced sex. Conclusions: Secondary school students in Nigeria from a polygamous family structure are more likely to have engaged in sexual activity than students from a monogamous family structure. This effect is partly explained by a higher likelihood of marriage during adolescence and forced sex. Students' sense of connectedness to their parents and school, regardless of family structure, decreases the likelihood of sexual activity, and fostering this sense may help reduce risky sexual behaviour among Nigerian youth.