Bovine fascioliasis causes important economic losses, estimated at COP$ 12,483 billion per year; its prevalence is 25% in dairy cattle. Parasitological techniques are required for it diagnosis. The Dennis technique, modified in 2002, is the one used in Colombia, but its sensitivity, specificity and validity are not known. To evaluate the validity and performance of the modified Dennis technique for diagnosis of bovine fascioliasis using as reference test the observation of parasites in the liver. We conducted a diagnostic evaluation study. We selected a convenience sample of discarded bovines sacrificed between March and June, 2013, in Frigocolanta for the study. We collected 25 g of feces from each animal and their liver and bile ducts were examined for Fasciola hepatica. The sensitivity, specificity, predictive positive value, predictive negative value, and validity index were calculated with 95% confidence intervals. The post-mortem evaluation was used as the gold standard. We analyzed 180 bovines. The sensitivity and specificity of the modified Dennis technique were 73.2% (95% CI=58.4% - 87.9%) and 84.2% (95% CI= 77.7% - 90.6%), respectively. The positive predictive value was 57.7% (95% CI= 43.3% - 72.1%) and the negative one 91.4% (95% CI= 86.2% - 96.6%). The prevalence of bovine fascioliasis was 22.8% (95% CI= 16.4% - 29.2%). The validity and the performance of the modified Dennis technique were higher than those of the traditional one, which makes it a good screening test for diagnosing fascioliasis for population and prevalence studies and during animal health campaigns.