Abstract Background Outbreaks of Chikungunya virus (CHIKV) occurred in Brazil during 2015–2017. Fortaleza was the city that reported the most cases. Methods The first round of a cohort study was conducted among women aged 15–39 y in Fortaleza, Brazil, in 2018 (Zika in Fortaleza). We collected sera to detect CHIKV IgG and IgM antibodies. Factors for CHIKV infection were identified using a Poisson regression model. Results We evaluated 1466 serum samples and 13.8% and 37.2% of women were found positive for CHIKV IgM and IgG antibodies, respectively. Living with more than four others in the same house and having an abandoned house nearby were associated with CHIKV infection. Being currently pregnant was associated with a decreased probability of CHIKV infection, which was also associated with pregnant women reporting using more repellent, both inside and outside the house, than non-pregnant women. Conclusions Crowding in households and abandoned houses nearby can increase potential transmission. Policies providing better living conditions and regulation of abandoned sites and buildings are necessary to control the mosquito population. Programmes providing repellant at low or no cost to pregnant women should be implemented in the neighbourhoods where arbovirus infections are endemic.