Our objective was to evaluate local guidelines regarding early steroid treatment in adult community acquired bacterial meningitis, and assess the actual treatment given and its correlation to clinical outcome. Patient outcome was obtained retrospectively from the medical records of 210 adults admitted to 47 hospitals in Denmark during 2002-2004 (population 5.4 million) and was combined with results from a questionnaire regarding treatment guidelines in these hospitals. In 36 of 47 departments responding to the questionnaire, 21 recommended early steroid treatment, but none did so initially during 2002. Early steroid treatment was given to 15% of patients and was given more often when recommended locally (41% vs 11%, OR=5.7 (2.4-13.5)). Unfavourable outcome was demonstrated rarely in patients treated with early steroids compared to the non-steroid group (17% vs 42%, p<0.05). In the 32 cases with petechial skin lesions, these were caused by pneumococci (15), meningococci (15), Staphylococcus aureus (1) and enterococci (1), and thus the presence of such lesions should not make the clinician abstain from early steroid treatment of bacterial meningitis. In conclusion, concordance with the new consensus of early steroid treatment was poor on a national basis, and better (41%) when adequate local guidelines were available. Early steroid treatment was associated with favourable outcome, and improved implementation of adequate guidelines may contribute to better patient outcome in bacterial meningitis.