Patients undergoing potentially painful interventional radiological procedures generally require a combination of analgesia and sedation. This sedation/analgesia should allow the patient to communicate while also remaining calm. Bispectral index (BIS) monitoring could be useful in achieving this. The primary end-point of our study was to compare the percentage time with optimal sedation, defined as Sedation Agitation Scale (SAS) grade 4, between a BIS-guided remifentanil/propofol regimen and a clinically guided regimen in 54 randomly allocated patients. The mean +/- sd percentage time with optimal sedation was significantly longer (P = 0.004) in the BIS group (76.6% +/- 14.7%) than in the SAS group (63.8% +/- 16.4%). There was a significant difference in the weighted mean infusion rates of remifentanil (P = 0.0067) and propofol (P = 0.0075) in the BIS group (0.066 +/- 0.027 microg.kg(-1) . min(-1) 1.59 +/- 0.44 mg.kg(-1) . h(-1)) compared with the SAS group (0.091 +/- 0.036 microg.kg(-1).min(-1) 1.92 +/- 0.43 mg.kg(-1).h(-1)), respectively. BIS values exhibited a temporal correlation to SAS scores (r2 = 0.72). In conclusion, a BIS-guided regimen was more effective than a SAS-guided regimen. The use of BIS resulted in fewer remifentanil and propofol doses. The targeted BIS range of 80-85 provided a sufficient functional level of sedation.