Background and aims: Per- and polyfluoroalkyl substances (PFAS) are non-degradable, man-made-chemicals with an elimination half-life of multiple years, causing accumulation in the environment and humans with potential harmful effects. However, longitudinal PFAS levels in human milk, daily PFAS intake and the association with infant plasma PFAS levels have never been reported. We investigated longitudinal PFOA and PFOS levels in human milk and the daily PFAS intake through infant feeding in the first 3 months of life, the most important determinants and the correlation with PFAS plasma levels at age 3 months and 2 years. Methods: In 372 healthy term-born Dutch infants, we determined PFOA and PFOS levels in human milk given at age 1 and 3 months, in 6 infant formula brands and in infant plasma at 3 months and 2 years, using liquid-chromatography-electrospray-ionization-tandem-mass-spectrometry(LC-ESI-MS/MS). We studied the associations between daily PFAS intake and predictive characteristics by multiple regression models. Results: PFOA and PFOS levels in human milk decreased between 1 and 3 months after delivery, regardless whether breastfeeding was given exclusively(EBF) or in combination with formula feeding. PFOA and PFOS could not be detected in any formula feeding. Daily PFAS intake(ng/kg) was highest in EBF-infants. Higher amount of human milk, older maternal age, lower parity and first-time breastfeeding were associated with higher daily intake. Daily PFAS intake in early life was strongly correlated with PFAS plasma levels at age 3 months and 2 years(R = 0.642–0.875, p