Elevated ratios of male to female births are emerging in Asian countries due to selective abortion of female foetuses. Little research has evaluated the possibility of sex selection among Asians in the West. We evaluated patterns in the secondary sex ratio (SSR, number of male per 100 female births) according to ethnicity in Québec, Canada. We performed a population-based retrospective analysis of live and still births in urban Québec, 1981-2004 (N = 1,612,614 live births). Outcomes included: (1) first and second order SSR over four consecutive 6-year intervals analysed according to parental mother tongue, parental birth region, and mothers still speaking her mother tongue at home, (2) adjusted relative risk (RR) of female birth by parental ethnicity, and (3) estimated number of unborn females. For the period 1987-1992, first order SSRs were elevated (range 118-121) for Indo-Pakistani parental mother tongue and mothers currently speaking Indo-Pakistani at home. Second order SSRs were not elevated. For Indo-Pakistani mother tongues, the RR of female birth was lower than French/English in the same period (adjusted RR 0.96, 95% CI 0.92-0.99). SSRs were not imbalanced in the late 1990s among Indo-Pakistani parents, or among Chinese ethnicity in any study period. The SSR in Québec was elevated in the 1980s and early 1990s among firstborn relative to subsequent born infants of Indo-Pakistani descent. The reason for this imbalance is unclear. Further research in other Western settings is necessary to evaluate the possibility of sex selection.