Aim: Anastomotic leakage is a serious complication after colorectal surgery. Different models have been studied in an attempt to detect anastomotic leak (AL) early, in order to reduce its morbidity and mortality. The objective of this study was to evaluate the clinical utility of the Dutch Leakage (DULK) and modified DULK scores to detect AL after elective colorectal surgery with colocolic or colorectal anastomosis. Methods: A total of 101 consecutive patients who underwent elective colorectal surgery with colocolic or colorectal anastomosis were retrospectively reviewed from May 2015 to October 2016. DULK and modified DULK scores were applied for each patient retrospectively. The main outcome measure is to match DULK score results with the clinical diagnosis or exclusion of an anastomotic leakage. Results: Of the 101 patients who underwent colorectal operations, eight had AL (7.9%), of which three had protective stomas. The DULK score had a sensitivity of 62.5%, specificity of 86.32%, positive predictive value of 27.78% and negative predictive value of 96.39%. Modified DULK score had a sensitivity of 75%, specificity of 76.43%, positive predictive value of 21.43% and negative predictive value of 97.26%. The DULK and modified DULK scores were able to detect AL on average 2 and 1.7 days, earlier than the clinical diagnosis, respectively. Conclusion: The DULK and modified DULK scores were an excellent tool for excluding AL after colorectal surgery. They are easily calculated and do not require sophisticated laboratory or radiological tests. A limitation is that this is a single‐centre study concentrated on evaluating a score with no long‐term outcomes. [ABSTRACT FROM AUTHOR]