Patient premedication for transesophageal echocardiography (TEE) is still under debate, especially the use of sedatives. We performed a double-blind, placebo-controlled study to clarify the efficacy of low-dose intravenous midazolam as premedication for TEE. Placebo or midazolam (< or =50 kg, 2.0 mg intravenously; 50 to 80 kg, 2.5 mg; and > or =80 kg, 3.0 mg intravenously) was given in addition to topical anesthesia to 200 consecutive in- and outpatients. Blood pressure, heart rate, and oxygen saturation were monitored. Patients were interviewed immediately, and 2 to 10 days after TEE. Sixteen patients received an additional dose of midazolam, and in 12 follow-up was incomplete. Patients taking midazolam reported less gag reflex at probe introduction and during TEE, as did the examiners (p < 0.05 to 0.0001). Probe manipulations were found to produce less discomfort after midazolam administration (p < 0.005). Midazolam patients experienced less dyspnea (p < 0.01) despite a minimal decrease in oxygen saturation of 2% (p < 0.0001). The following day patients taking midazolam reported less sore throats, and painful swallowing was less frequent (p < 0.01 to 0.001). Systolic blood pressure decreased slightly in the midazolam group (132 +/- 24 to 121 +/- 20 mm Hg, p < 0.0001). The rate of minor complications showed no difference. Thus, TEE probe introduction and manipulation was tolerated better after low-dose midazolam premedication, and patients experienced less pharyngeal discomfort the day after. Midazolam was well tolerated and the complication rate did not increase. Thus low-dose, short-acting benzodiazepine premedication improves patient comfort during and after TEE and generous use can be recommended.