Objectives Data regarding women infected with SARS-CoV-2 during early trimesters are scarce. We aimed to assess preterm birth (PTB) and small-for-gestational-age (SGA) rates in a large and unselected cohort by trimester at infection and overall. Design A retrospective cohort study including all women with a positive SARS-CoV-2 RT-PCR test during a non-ectopic singleton pregnancy between February 21st 2020 and July 2nd 2021 (N = 2753). Each infected woman was matched to a non-infected pregnant woman by age, last menstruation date, sector, and socioeconomic status. Methods Logistic regression was conducted to assess the risks of PTB and SGA including an interaction between group and trimester of infection. Multivariable models included underlying diseases, previous abortions and null parity. Subgroup analyses were conducted on symptomatic infected women and matched non-infected women. Results A total of 2753 /2789 (98.7%) eligible women that were infected during pregnancy could be matched, among them, 17.4% and 48.4% were infected during the first and third trimesters, respectively. While first and second trimester infections were not associated with PTB (p>0.8), third trimester infections and in particular after 34 weeks of gestation had a greater risk of PTB with adjusted ORs of 2.76 (95% CI 1.63–4.67) and 7.10 (95% CI 2.44–20.61), respectively. PTB risk was further heightened in symptomatic third trimester infections (OR = 4.28, 95% CI 1.94–9.25). SGA risk was comparable between study groups across all trimesters of infection. Pregnancy loss incidence was similar in both groups (adjusted OR = 1.16; 95% CI 0.90–1.50). Conclusion SARS-CoV-2 infection was associated with increased risk of PTB only among women infected during late pregnancy, particularly among symptomatic women.