Introduction: Neonates admitted to Neonatal Intensive Care Units (NICU) have a substantial risk for healthcare-associated infections (HAI). A high incidence of HAI was found in Maternidade Júlio Dinis (MJD) NICU and led to changes in daily clinical practice. Objectives: To evaluate the effectiveness of new strategies for prevention of infection implemented in the NICU of MJD. Methods: We conducted a prospective study that included all infants admitted to the NICU of MJD during the 16 months before (group 1) and 16 months after (group 2) the protocol change, comparing the rates of infection in these two time periods. Results: 528 infants were included in group 1 and 593 in group 2, with no statistical evidence of different characteristics between them (weight, gestational age, CVC utilization rate, duration of parenteral nutrition and invasive ventilation). Overall mean infection indices decreased between the two periods: incidence -density of infection – sepsis, pneumonia, necrotizing enterocolitis and meningitis – (10,0 vs. 6,3; p=0,0007), incidence- -density of clinical plus laboratory -confirmed sepsis (9,6 vs. 5,9; p=0,0007), incidence -density of laboratory -confirmed sepsis (6,8 vs. 4,6; p=0,015), incidence rate of central vascular catheter (CVC) -associated sepsis (30,3 vs. 22,4; p=0,021) and coagulase- -negative Staphylococcus isolation rate (85,7% vs. 71,1%; p=0,041). Although incidence rate of laboratory -confirmed CVC- -associated sepsis decreased (22 vs. 19), the difference was not significant. Similar tendencies were observed considering very low birth weight neonates. Conclusions: This study demonstrated that an evidence- -based intervention was effective, resulting in a decrease of 37% of the HAI in the NICU of MJD.