The objective of this paper is to compare the accuracy of standard, non-contrast endosonography (EAS) with contrast-enhanced endosonography (CE-EAS) in the diagnosis of anal fistulas. The group consisted of 126 patients (mean age: 43.1 years) with the clinical diagnosis of anal fistula. For anal endosonography, a Bruel & Kjaer unit with a 7.0-MHz transducer was used with a 3% solution of hydrogen peroxide as the contrast agent (1, 2, 3). In each case, EAS and CE-EAS diagnoses of the type and complexity of anal fistula, as well as the location of the internal opening, were determined. Results showed that CE-EAS was significantly more accurate in diagnosing the type of anal fistulas than NC-EAS (97 vs. 94%, respectively; P=02275), and in differentiating simple from complex tracks (92 vs. 75%, respectively; P<0.00001). CE-EAS was much more accurate in patients with recurrent fistulas (57 vs. 92%, respectively; P<0.00006), whereas in a subgroup of primary tracks, both methods were of comparable accuracy. Sensitivities of CE-EAS and EAS for internal opening were 89 and 65%, respectively. The conclusion of this paper is that CE-EAS significantly increases the accuracy of standard non-contrast EAS and is especially beneficial for the differentiation between simple and complex tracks.