This study aimed to determine the relationships between prenatal PM2.5 exposure and childhood growth trajectories during the first 6 years of life. A total of 47,625 pairs of mothers and children were recruited from a prospective birth cohort conducted between 2011 and 2013 in Wuhan, China, and followed for 6 years. We used the group-based trajectory models to classify the population into three trajectory groups: slow growth (n = 13,671, 28.7%), normal growth (n = 29,736, 62.4%), and rapid growth (n = 4218, 8.9%). Multinomial logistic regression models were used to determine the associations of prenatal PM2.5 exposure and childhood growth trajectories. Compared to normal growth trajectory, increased PM2.5 exposure in trimester 1, trimester 2 and the entire pregnancy showed significant associations with an increased risk of the slow growth trajectory but reduced the risk for the rapid growth trajectory, significant association of prenatal PM2.5 exposure with rapid growth trajectory was only observed in the trimester 3. Stratified analyses displayed relatively stronger associations among those mothers with maternal age over 35 years, pre-pregnancy BMI ≥ 25 kg/m2, and previous delivery experience. Prenatal exposure to PM2.5, particularly during the midpoint period of pregnancy, was more likely to have a slow growth trajectory and a lower risk of rapid growth trajectory. Maternal age, pre-pregnancy BMI, and previous delivery experience might modify these associations.