Seventy-eight patients with active Crohn's disease participated in a randomized, double-blind, cross-over trial. The study comprised two 4-mo period. The purpose was to test the efficacy of metronidazole in comparison with that of sulfasalazine. As the main evaluation criteria the Crohn's Disease Activity Index and plasma levels of orosomucoid were chosen. In the first period no difference in efficacy as measured by Crohn's Disease Activity Index was found between the treatment groups. The reduction of the plasma orosomucoid level was significantly more pronounced in the metronidazole group. The hemoglobin concentration increased more in this group than in the sulfasalazine group, possibly due to a toxic effect of sulfasalazine. The erythrocyte sedimentation rate decreased similarly with both drugs. In 15 patients who had active disease throughout the first period, Crohn's Disease Activity Index decreased significantly in the second period for those who switched to metronidazole, but not for those who switched to sulfasalazine. After crossover, no apparent further change in Crohn's Disease Activity Index occurred in either of the treatment groups among patients who had responded favorably in the first period. The plasma concentration of orosomucoid increased significantly among the patients in the sulfasalazine group but not in the metronidazole group. It is therefore concluded that metronidazole is slightly more effective than sulfasalazine in the treatment of crohn's disease. It is worthwhile switching the drug regimen from sulfasalazine, when it fails, to metronidazole, but not from metronidazole to sulfasalazine.