The levels of arachidonic (AA) and docosahexaenoic acid (DHA) fall in the first days of life of preterm infants and it was hypothesised that this could relate to poor neurological and physiological outcomes. We explored the fatty acid changes in over 1000 very preterm infants enrolled in the N3RO trial (1) at two timepoints, namely at enrolment and at the end of oral supplementation period (36 weeks’). Infants were enrolled within the first week of life (mean age 3.3+/-1.5 days of life). Infants (n = 1077) were randomised to either enteral emulsion intervention with 60 mg/kg/d of DHA or no DHA soy oil. Most infants received small, graded volumes of enteral human milk or formula, as soon as possible after birth, and were administered one of 3 intravenous fats (at 1–3 g/kg/d) in the first week of life, according to unit protocols, that varied in the proportion of linoleic and oleic acids. 96.7% of infants received human milk. At enrolment, considerable variation was observed in the proportions of blood AA (1.4%–18.0%) and DHA (0.4%–5.7%) concentrations. DHA was not associated with gestational age at birth, but AA was lower by 0.11% for each week of gestation (95% CI: 0.07%, 0.15%). There was an inverse relationship between age of sampling after birth and DHA (0.07% lower per day; 95% CI: 0.04%, 0.09%) and AA (0.09% lower; 95% CI: 0.04%, 0.14%). Preterm infants whose blood was sampled later after birth had lower DHA and AA status relative to those sampled earlier after birth and were higher in linoleic acid and oleic acid, which are the major components of intravenous lipid emulsions. Following high-DHA supplementation, mean DHA of intervention group infants increased to 3.86%, while it decreased to 2.53% in the control group (mean difference 1.33%, 95% CI 1.27 to 1.40, p < 0.0001). AA was lower at 36 weeks’ than at enrolment, with the decrease marginally more pronounced in the intervention group (8.11%) compared with the control group (8.84%, mean difference -0.71%, 95% CI -0.86 to -0.57, p < 0.0001). All fatty acids regressed towards a new diet-dependent mean at 36 weeks regardless of the initial level. DHA supplementation successfully prevented the typical decline in DHA and modestly exacerbated the decline in AA. The results demonstrate the major influence both intravenous and oral nutrition sources have on the levels of AA and DHA of critically ill infants and the need to more completely understand whether intravenous and oral feeds need to be realigned with the fatty acid composition of infant blood at birth.