Background and aims: Per- and polyfluoroalkyl substances (PFAS) are a potential hazard for public health. These man-made-chemicals are non-degradable with an elimination half-life of multiple years, causing accumulation in the environment and humans. Rodent studies demonstrated that PFAS are harmful, especially when present during the critical window in the first months of life. Because longitudinal data during infancy are limited, we investigated longitudinal plasma levels in infants aged 3 months and 2 years and its most important determinants. Methods: In 369 healthy term-born Dutch infants, we determined plasma PFOS, PFOA, PFHxS, PFNA and PFDA levels at age 3 months and 2 years, using liquid chromatography-electrospray-ionization-tandem-mass-spectrometry (LC-ESI-MS/MS). We studied the associations with maternal and child characteristics by multiple regression models. Results: At age 3 months, median plasma levels of PFOS, PFOA, PFHxS, PFNA and PFDA were 1.48, 2.40, 0.43, 0.23 and 0.07 ng/mL, resp. Levels decreased slightly until age 2 years to 1.30, 1.81, 0.40, 0.21 and 0.08 ng/mL, resp. Maternal age, first born, Caucasian ethnicity and exclusive breastfeeding were associated with higher infant's plasma levels at age 3 months. Levels at 3 months were the most important predictor for PFAS levels at age 2 years. Infants with exclusive breastfeeding during the first 3 months of life (EBF) had 2–3 fold higher levels throughout infancy compared to infants with exclusive formula feeding (EFF), with PFOA levels at 3 months 3.72 ng/mL versus 1.26 ng/mL and at 2 years 3.15 ng/mL versus 1.22 ng/mL, respectively. Conclusion: Plasma PFAS levels decreased only slightly during infancy. Higher levels at age 3 months were found in Caucasian, first-born infants from older mothers and throughout infancy in EBF-infants. Our findings indicate that trans-placental transmission and breastfeeding are the most important determinants of PFAS exposure in early life.