Background: Hepatic dysfunction is a complication associated with parenteral nutrition (PN). Our primary objective was to study the relationship between doses of intravenous fish oil (FO) emulsion in PN and the variation in the main liver function tests (LFTs) in hospitalized PN-treated adults. As a secondary objective, we studied the safety of FO administration. Methods: We conducted a retrospective study in adult patients receiving FO supplementation in PN. Demographic, nutritional and safety variables were collected. Variation of LFTs was defined as the difference between values just before the first administration of FO and values at the end of PN. A multiple linear regression was performed to study the association between PN-lipids (FO or vegetable) and the variation of each LFT; the following variables were used to adjust the effect of lipids: sepsis, length of stay in the intensive care unit and lipids dose. Student t-test was used to study safety variables. Data were analyzed using SPSS 19.0. Results: Patients (53, median age 68 years (24-90); 62 % men) with the principal diagnosis of digestive neoplasm (42 %) received PN for a median of 19 (7-75) days. In the multivariate analysis, the amount of FO was related to a decrease in gamma-glutamyl transferase (GGT) (B = -2.23;CI95 % = -4.41/-0.05), a decrease in alkaline phosphatase (AP) (B = -1.23;CI95 % = -2.07/-0.37), and a decrease in alanine aminotransferase (ALT) (B = -0.82; CI95 % = -1.19/-0.44). No differences were found in safety variables. Conclusions: GGT, AP and ALT improved with FO PN-supplementation. Moreover, the improvement was greater when the doses of FO were higher. FO administration in PN is safe. [ABSTRACT FROM AUTHOR]