In this study we review the rationale for using selective digestive decontamination (SDD) in critically ill patients, and its effects on clinical outcomes and rates of infection with antimicrobial-resistant microorganisms. SDD consists of the application of nonabsorbable antibiotics to the oropharynx and through a nasogastric or nasoenteral tube, in association with a 4-day course of an intravenous third-generation cephalosporin. The enteral component aims at preventing oral and rectal colonization with potentially pathogenic nosocomial aerobic gram-negative bacilli and yeasts while preserving normal protective anaerobic enteral flora. The short-course systemic component aims at eradicating oral endogenous gram-positive bacteria. SDD decreases the risk of nosocomial infections, and reduces by one-quarter the mortality of patients on mechanical ventilation in settings with low prevalence of antibiotic resistance. Evidence from randomized trials suggests that SDD does not increase rates of antimicrobial-resistant microorganisms, and may reduce resistance rates to some antibiotics. However, several limitations decrease our confidence on these data, particularly for settings with high baseline rates of antimicrobial-resistant microorganisms. Although SDD has a clear potential to improve clinical outcomes of critically patients, its long-term ecologic effects on rates of antimicrobial resistant require appropriate assessment by large multinational cluster randomized trials. Before these results are available, the use of SDD cannot be recommended in most parts of the world, except in settings with very low baseline prevalence of antibiotic resistance.