In recent years, rainfall induced ‘urban storm water-logging’ (USWL) events are experiencing in Chittagong city like other urbanized parts of Bangladesh. To mitigate this there is an urgent need to predict the USWL beforehand and a numerical model could help. Thus, this paper aimed to use a hydrological model i.e. HEC-HMS through field survey during 2013–2014 and a questionnaire survey. With the support of secondary data source viz. daily newspaper, intensive questionnaire survey and field visits identified 13 most vulnerable USWL locations and based on these a USWL depth–duration–frequency curve was developed. This showed during water logging (i.e. May–July), the depth of logged water rises 0.3–1.27 m causing adjacent dwellers sufferings upto 13 times per year and the inundation period is up to 48 h. Using these, HEC-HMS model was setup and the correlation obtained with the field measurement as R2 values 0.83 and 0.77 during calibration and verification period respectively. Once a real time hydrological dataset is available, the validated model supposed to provide useful information in the decision support system.