Objective To explore the effects of perinatal infection on retinopathy of prematurity(ROP). Methods A retrospective cohort study was performed to analyzed the clinical data of 238 preterm infants at gestational age ≤32 weeks who were delivered in Guangdong Women and Children Hospital from November 2014 to October 2015 and ROP screening examinations.Observation was not terminated until they were 45 weeks of corrected gestational age.Mild ROP was defined as having stage 1 or stage 2 ROP in zone Ⅱ or Ⅲ without additional disease, and severe ROP was defined as stage 3 or higher, any ROP in zone Ⅰ, prethreshold/threshold, with additional disease, and aggressive posterior retinopathy of prematurity (AP-ROP). Medical records of eligible preterm infants were retrospectively reviewed and analyzed.Occurrences of ROP, severe ROP, and clinically significant ROP requiring surgical treatment were assessed. Results The mean gestational age of the cohort was (30.10±1.34) weeks (25.29-32.00 weeks) and the mean birth weight was (1 373±272) g(720~2 330 g). ROP was diagnosed in 76 of 238 infants (31.9%), including 39 cases with mild ROP (16.4%) and 37 cases with severe ROP (15.5%). Surgical treatment was performed on 22 infants (9.2%). In the patients with ROP, the time to develop ROP from birth was (35.16±14.26) d and the mean time of its most serious stage was (44.62±18.99) d. In 22 patients with ROP who required surgical treatment, the time of surgical treatment was (50.27±17.24) d. In univariate analysis, maternal perinatal infection disease was found to be associated with ROP occurrence (χ2=7.891, P=0.005) and ROP progression requiring surgical treatment (χ2=4.494, P=0.034 ). Small gestational age, low birth weight and long-term oxygen therapy were found to be associated with ROP occurrence and severe ROP ( gestational age: t=-5.803, P<0.001; t=-5.290, P<0.001; t=-4.150, P<0.001; birth weight: t=-4.942, P<0.001; t=-4.058, P<0.001; t=-3.126, P=0.002; the duration of oxygen therapy: t=2.351, P=0.020; t=2.473, P=0.018). Apgar scores ≤7 at 1 min and 5 min were found to be associated with severe ROP ( χ2=4.803, P=0.028 ). Neonatal sepsis and neonatal fungal infection were found to be associated with ROP occurrence ( χ2=6.071, P=0.014; χ2=4.070, P=0.044). Neonatal fungal infection was also found to be associated with severe ROP ( χ2=5.479, P=0.019 ). Multivariate regression analysis indicated that maternal perinatal infection disease was associated with an increased risk of ROP and ROP progression requiring surgical treatment (OR= 2.837, P=0.023; OR=4.087, P=0.012). Maternal preeclampsia was also associated with an increased risk of ROP (OR= 2.506, P=0.040). Gestational age was an important risk factor for the development of ROP.The smaller the gestational age was, the higher the rate of occurring ROP and severe ROP (OR=0.518, 0.508, 0.520, all P<0.001). Conclusions Both fetal and neonatal exposure to infection appear to contribute to the increase of ROP risk in the preterm infants at gestational age ≤32 weeks.Maternal perinatal infection disease and maternal preeclampsia were independently associated with ROP occurrence and ROP progression in the preterm infants at gestational age ≤32 weeks. Key words: Prenatal factor; Inflammation and infection; Retinopathy of prematurity; Retrospective cohort study