'Low-dose' dopamine is frequently used in intensive care units (ICU) for its presumed renoprotective effects, but prospective and retrospective studies have so far not proven prevention or amelioration of renal injury. Data on renal perfusion following dopamine infusion are limited. In order to circumvent the problem of patient heterogeneity in the ICU setting, we used a crossover design in a prospective, double-blind randomized controlled study to investigate the effect of 'low-dose' dopamine on renal resistance indices, as determined by Doppler ultrasound. Forty patients, 10 without and 30 with acute renal failure (ARF, defined as doubling of baseline creatinine or an increase above 2 mg/dl), were included. Dopamine (2 mug/kg min) or placebo was given intravenously in alternating sequence for four subsequent periods of 60 min, starting randomly with either dopamine or placebo. Resistive (RI) and pulsatility index (PI) were closely correlated, positively related to serum creatinine values at baseline and highly reproducible during the two paired infusion periods. Dopamine reduced renal vascular resistance in patients without ARF (median RI/PI from 0.70 to 0.65/1.20 to 1.07, P<0.01) but increased resistance indices in patients with ARF (median RI/PI from 0.77 to 0.81/1.64 to 1.79, P<0.01) in the absence of effects on systemic hemodynamics. Subgroup analysis of patients with ARF revealed that dopamine induced renal vasoconstriction above 55 years (n=22) and in patients not receiving norepinephrine (n=20). In conclusion 'low-dose' dopamine can worsen renal perfusion in patients with ARF, which adds to the rationale for abandoning the routine use of 'low-dose' dopamine in critically ill patients.