Gentamicin is often used to treat listeriosis, particularly in patients with meningitis; nonetheless, some clinicians question this practice because of the drug's associated nephrotoxicity and inability to cross the blood-brain barrier. The aim of this study was to evaluate predictors of mortality and the impact of aminoglycosides on outcome in patients with listeriosis. We conducted a retrospective study of all non-pregnant adult patients with Listeria monocytogenes infection detected in sterile body fluids between 1983 and 2006. Early mortality was defined as death occurring between days 3 and 14 after admission, and late mortality as in-hospital death after 14 days. Of 118 episodes, 16 were excluded because patients died in the first 48 h. Among the 102 patients analysed, 33 (32%) had received combined beta-lactam and aminoglycoside therapy and 69 (68%) beta-lactam monotherapy. Both groups had similar demographic and clinical features, and rate of appropriate initial therapy. Overall mortality was 21/102 (20.6%). Early overall mortality was 11.8%: 27.3% (9/33) in the combined group and 4.3% (3/69) in the monotherapy group (P = 0.003). Late mortality was 8.8%. In the multivariate analysis, the factors predicting early mortality were renal failure, previous corticosteroid therapy and age >65 years, whereas neoplastic disease and coma were associated with late mortality. Gentamicin administration did not decrease early mortality, but seemed to increase it. In the late mortality analysis, gentamicin use had no impact. In an analysis with the propensity score method for the use of aminoglycosides, combined therapy with this antibiotic was associated with an increasing trend for early mortality (OR 3.40, 95% CI 0.82-14.07). The addition of aminoglycosides to treatment for listeriosis did not improve the patients' outcome.