The decision on the right empirical treatment in bacteremia places particular demands on the clinician. As long as no microbiological diagnosis can be immediately drawn, the clinical diagnosis together with knowledge of local antimicrobial resistance must determine the antimicrobial choice. The use of several amplification, hybridization, and mass spectrometry methods has been studied in patient cohorts in comparison with blood culture-based conventional techniques. However, no clinical outcome trials have been conducted in which the use of these novel methods would guide antimicrobial therapy. Local differences in bacterial antimicrobial resistance cause differences in the regional need for molecular methods for the early detection of resistance mechanisms. The implementation of novel methods in clinical use requires active discussion between laboratory experts and clinicians. Providing rapid susceptibility results using conventional methods can lead to timely changes to appropriate antimicrobial therapy and the costs are lower than with the molecular methods. Gram-stain information in combination with clinical data is an underestimated, underused, rapid, and economical means of assessing the etiology of blood stream infection.