Background  Transfusion of allogeneic red blood cells (RBC) may be associated with side effects. This study aimed to assess whether an association could be detected between transfusion practice and the occurrence of complications after laparotomy. Study design and methods  This study is an observational analysis of data from a randomized trial in 1400 patients who underwent laparotomy. A subgroup of 224 transfused patients with an intraoperative blood loss ≥200 ml were included in the analysis. Logistic regression analysis was used to investigate risk factors for postoperative complications. The ratio of intraoperative RBC transfusion to blood loss was computed, and patients grouped by the median into a liberal transfusion practice (ratio equal to or above the median) and a restrictive transfusion practice group (ratio below the median). Results  Surgical site infection occurred in 27% of patients in the liberal group vs. 20% of patients in the restrictive group with an OR of 1·5 [95% CI: 0·8-2·9] (P = 0·18) and an OR of 1·2 [95% CI: 0·5-2·9] (P = 0·73) when adjusting for known confounding variables. Pneumonia occurred in 14% vs. 8% in the liberal and restrictive group, respectively (adjusted P = 0·07), and admission to the intensive care unit was 15% vs. 7%, respectively (adjusted P = 0·02), but no other significant differences were found. Conclusion  A liberal transfusion practice was not significantly associated with postoperative complications, but pneumonia tended to be more common in the liberal group, which was more often admitted to the intensive care unit.