Objective To investigate the effect of L-carnitine on lipid and bilirubin metabolism in extremely premature infants (EPI). Methods The clinical data of preterm infants with gestational age <28 weeks hospitalized from August 2016 to December 2019 were retrospectively analyzed. Preterm infants were divided into L-carnitine group and non-L-carnitine group according to the prophylactic use of L-carnitine within 72 hours after admission, and the levels of lipids and bilirubin were compared between the two groups. Results A total of 261 EPI (151 boys and 110 girls) were enrolled, and the mean gestational age was (26.5±1.3) weeks and the mean birth weight was (907.3±175.9) g. There were 139 patients in L-carnitine group and 122 in non-L-carnitine group. The 1-minute Apgar score and the proportion of prenatal dexamethasone use in L-carnitine group were higher than those in non-L-carnitine group, and the differences were statistically significant (P<0.05). Compared with non-L-carnitine group, L-carnitine group had lower total bilirubin level at week 1 and 2, higher bile acid level at week 2, lower direct bilirubin level at week 4, higher triglyceride level at week 1, lower blood urea nitrogen and creatinine levels and higher alanine aminotransferase level at week 1 and 2, and the differences were statistically significant (P<0.05). Among 261 patients, 66 had parenteral nutrition associated cholestasis (PNAC), including 33 patients (23.7%) in the L-carnitine group and 33 (27.0%) in the non-L-carnitine group. Binary logistic regression analysis showed that triacylglycerol level at the fourth week of hospitalization was a risk factor for PNAC (P<0.05). There was no significant correlation between prophylactic use of L-carnitine and the occurrence of PNAC (P>0.05). Conclusions Early prophylactic use of L-carnitine in EPI can regulate bilirubin and lipid metabolism, but it will not reduce the incidence of PNAC, nor affect liver and kidney function.