Although 30-50% of constipated patients exhibit dyssynergia, an optimal method of diagnosis is unclear. Recently, consensus criteria have been proposed but their utility is unknown. To examine the diagnostic yield of colorectal tests, reproducibility of manometry and utility of Rome II criteria. A total of 100 patients with difficult defecation were prospectively evaluated with anorectal manometry, balloon expulsion, colonic transit and defecography. Fifty-three patients had repeat manometry. During attempted defecation, 30 showed normal and 70 one of three abnormal manometric patterns. Forty-six patients fulfilled Rome criteria and showed paradoxical anal contraction (type I) or impaired anal relaxation (type III) with adequate propulsion. However, 24 (34%) showed impaired propulsion (type II). Forty-five (64%) had slow transit, 42 (60%) impaired balloon expulsion and 26 (37%) abnormal defecography. Defecography provided no additional discriminant utility. Evidence of dyssynergia was reproducible in 51 of 53 patients. Symptoms alone could not differentiate dyssynergic subtypes or patients. Dyssynergic patients exhibited three patterns that were reproducible: paradoxical contraction, impaired propulsion and impaired relaxation. Although useful, Rome II criteria may be insufficient to identify or subclassify dyssynergic defecation. Symptoms together with abnormal manometry, abnormal balloon expulsion or colonic marker retention are necessary to optimally identify patients with difficult defecation.