The effect of midazolam and vecuronium on mean arterial pressure (MAP) and mean cerebral blood flow velocity (MCBFV) was evaluated in premature infants (birthweight 550-2,560 g; gestational age 26-36 weeks) randomised to receive either 0.1 mg/kg midazolam (n = 7) or 0.05 mg/kg vecuronium (n = 8) intravenously. MAP, by means of an indwelling arterial catheter, and MCBFV, by means of non-invasive pulsed-Doppler of the middle cerebral artery, were measured every 5 min, starting at 10 min prior to until 1 h after drug administration. A transient 25-43% decrease in MCBFV (mean 0.06 m/s) dependent on a 8-23% decrease in blood pressure (mean 9 mm Hg) was noted in all patients within 15 min following administration of midazolam, which returned to baseline values within 1 h. In 2 out of 7 infants, a plasma expander was required. In contrast, vecuronium only decreased the MCBFV in 3 of 8 infants. Thus, a bolus of midazolam transiently decreased blood pressure and MCBFV, and should be used cautiously in sick preterm infants.