OBJECTIVE It has been proposed recently that red blood cell transfusions (RBCT) might increase severity in infants with necrotizing enterocolitis (NEC). We intend to study if patients who have received red blood transfusions before the onset of NEC develop more severe forms of this entity. METHODS A retrospective study was carried out including prematures with NEC. Two groups were considered: with previous RBCT (TR) and without previous RBCT (No-TR). The main outcomes of the study were severity of NEC, according to the Bell stage (BS), surgical treatment and mortality. Patients who were treated with RBCT 48 hours prior to the onset of NEC symptoms were analysed separately afterwards. Comparison of groups was made with the Fisher test or the Chi square test for the BS, surgery, mortality and nominal variables; the U Mann-Whitney test was used for numeric variables. RESULTS Forty-six patients were included for the investigation (28 in TR and 18 in No-TR). In the TR Group 20/28 neonates reached a BS II; 8/28 BS III; 10 were operated on and there were 7 deaths. In the No-TR group 14/18 patients were classified as EB II; 4/18 as BS III, 3 patients needed surgery and Idied. No relation was found between RBCT and BS (p = 0.39), RBCT and surgery (RR = 2.7; CI 95%: 0.64-11.97; p = 0.19), or RBCT and mortality (RR = 4.5; CI 95%: 0.6-36.6; p = 0.09). In those patients who received a RBCT 48 hours before the initial symptoms there were 3 EB II and 3 EB III; 4 infants required surgical treatment and there were 2 exitus. Comparing this subgroup and the rest of the sample there were only significant differences in the need of surgical treatment: patients who received a RBCT 48 hours before the onset of NEC were at an increased risk for surgery of 2.6 (CI 95%: 1.2-5.1; p = 0.045) but there were not different when surgical treatment and mortality were considered. CONCLUSIONS These results do not support clearly the hypothesis that there is a relation between previous treatment with RBCT and the severity of NEC.