Background: Preoperative anaemia and red blood cell (RBC) transfusions are associated with poorer clinical outcomes. It is unknown whether perioperative RBC transfusions mediate the relationship between preoperative haemoglobin levels and postoperative outcomes. Methods: This was a prospective observational study among patients aged >= 50 yr undergoing elective major noncardiac surgery from four Swedish hospitals. The co-primary outcomes were 1-yr major adverse cardiovascular and cerebrovascular events (MACCEs) and all-cause mortality. The secondary outcome was a composite of 30-day mortality, MACCEs, acute kidney injury (AKI), pulmonary embolism, anastomotic leak, and postoperative infection. Mediation analyses were conducted with preoperative haemoglobin as the exposure and RBC transfusion as a mediator. Results: Among 1060 patients (mean age 70 [SD 9] yr; 472 [45%] women), 171 patients (16.1%) developed 1-yr MACCEs, and 105 patients (9.9%) died within 1 yr. Preoperative haemoglobin levels were significantly associated with both 1-yr MACCEs (b=-0.015, P=0.041) and all-cause mortality (b=-0.028, P<0.001). Volume of RBC transfusion was not directly associated with the outcomes and did not mediate the relationship between preoperative haemoglobin levels and 1-yr MACCEs (b=-0.001, P=0.451) or all-cause mortality (b=-0.002, P=0.293). For the secondary outcome, RBC transfusions had a significant mediating effect between preoperative haemoglobin and the composite 30-day outcome; however, no direct association was observed (b=0.006, P=0.554). Conclusions:<bold> </bold>Preoperative haemoglobin levels were significantly associated with 1-yr MACCEs and all-cause mortality. This effect was not mediated by perioperative RBC transfusions. Further research is needed to confirm these findings.