Summary: In order to quantify the impact of therapeutic drug monitoring (TDM) services on the average number of serum drug assays (SDAs) ordered for patients, a meta-analysis was conducted on the data collected from 16 published and unpublished studies. Overall, TDM has no effect on the number of SDAs ordered per patient. However, there were a number of moderator variables that influenced the results. In studies utilizing control groups for comparison, TDM service-monitored patients were ordered fewer SDAs than nonmonitored patients. When data from the same institution were collected before and after implementation of a TDM service, there was an increase in the number of SDAs per patient. Aminoglycoside patients monitored by a TDM service were ordered more SDAs and nonaminoglycoside monitored patients were ordered fewer SDAs than nonmonitored patients. When the pharmacy department determined when the SDAs were to be collected and adjusted the dose, monitored patients were ordered fewer SDAs than nonmonitored patients. However, when the pharmacy department did not have control of these parameters, the TDM patients were ordered more SDAs than nonmonitored patients. Methodologic limitations of the studies and their possible effects on the results are discussed.