PURPOSE:The purpose of this study was to evaluate the use and timing of computed tomography in the treatment of patients with acute left‐sided diverticulitis.METHODS:We reviewed our four‐year experience of 47 patients with the diagnosis of acute diverticulitis. We have evaluated the benefits of admissionvs.delayed computed tomography in patients with this diagnosis.RESULTS:Of the 47 patients, 17 were diagnosed on clinical grounds alone, treated, and released. Thirty patients had their clinical diagnoses of diverticulitis evaluated with either computed tomographic scan (26) or laparotomy (4). Eleven of those 30 (36 percent) patients were found to have normal computed tomographic scans, indicating inaccurate clinical diagnosis, and all patients who underwent laparotomy had the pathologic diagnosis of diverticulitis. Six of the 47 patients had abscesses, but only 2 were identified at the time of admission. The remaining four abscesses were identified on delayed computed tomographic scans after failure of medical therapy. Thirty‐seven hospital days were used by patients with inaccurate diagnoses before their computed tomographic scans. Analysis of cost revealed that a computed tomographic scan for all 47 patients would have cost less than the expense of admission for just the 11 patients who had normal computed tomographic scans.CONCLUSION:Routine admission computed tomographic scan for patients with acute diverticulitis leads to more accurate diagnosis, earlier identification of complications, and possible decreased hospital costs.