A total of 686 adult patients with complicated urinary tract infections were enrolled in a double-blind, randomised, multicentre study to compare sparfloxacin (200 mg loading dose on day 1 followed by 100 mg daily) with ciprofloxacin (500 mg orally twice daily) for 10 to 14 days. Urinary tract infection was defined as pyuria and bacteriuria (cfu > or = 10(5)/mL). Evaluations were performed at four time-points. The clinical efficacy of the two antibacterial agents was equivalent at the end of treatment: clinical cure in 88.6% of the intent-to-treat population and 87.3% in the evaluable population treated with sparfloxacin compared to 85.4% and 84.8% of the intent-to-treat and evaluable populations, respectively, treated with ciprofloxacin. The clinical results were also equivalent at follow-up. The bacteriological efficacy of the two agents was not equivalent. At the end of treatment, bacteriological cure was observed in 72.6% of the intent-to-treat and 72.1% of the evaluable populations treated with sparfloxacin and in 81.4% and 80.8% of the intent-to-treat and evaluable populations, respectively, treated with ciprofloxacin. The difference was primarily because of a higher number of persisting pathogens, which included Enterobacteriaceae other than Escherichia coli, Pseudomonas aeruginosa and enterococci, which exhibited moderate susceptibility to sparfloxacin. Tolerability was similar in the two treatment groups.