The benefits of a pharmacy-based, active therapeutic drug monitoring (TDM) service (ATM) on outcomes were examined in a prospective study at four hospitals. ATM involved pharmacokinetic dosage optimization at the start of treatment, subsequent Bayesian adaptive control, and frequent patient evaluation. Cost-effectiveness was calculated based on real costs. The ATM group comprised 105 patients and 127 patients with nonguided TDM who were followed up as controls. Forty-eight of the ATM and 62 of the nonguided TDM patients had an infection on admission. Peak concentrations in ATM patients were significantly higher (10.6+/-2.9 mg/L; nonguided TDM, 7.6+/-2.2 mg/L; p < 0.01). Trough levels in the ATM group were significantly lower (p < 0.01). There was a trend toward lower mortality in the ATM group (nine of 105 versus 18 of 127; p = 0.26) that was significant for patients with an infection on admission (one of the 48 ATM patients died versus nine of the 62 nonguided TDM patients; p = 0.023). ATM reduced the length of hospital stay for all patients in the study (20.0+/-1.4 days; nonguided TDM, 26.3+/-2.9 days; p = 0.045) and for patients admitted with an infection (12.6+/-0.8 days; nonguided TDM, 18.0+/-1.4; p < 0.001). The incidence of nephrotoxicity was reduced from 13.4% (nonguided TDM) to 2.9% (p < 0.01). With ATM, total costs were lower for all patients (Dutch guilders [DFL], 13,125+/-9,267; nonguided TDM, DFL 16,862+/-17,721; p < 0.05) and for patients admitted with an infection (DFL 8,883+/-3,778; nonguided TDM, DFL 11,743+/-7,437; p < 0.01). Goal-oriented, model-based dosing of aminoglycosides resulted in higher antibiotic efficacy, shorter hospitalization, and reduced incidence of nephrotoxicity. By combining efficacy with savings, ATM offered a significant alternative to usual care.