Management of reservoir water quality is a global challenge due to the natural process of eutrophication and anthropogenic aggravation. In Sri Lanka, irrigation reservoirs support several secondary uses such as fish production, livestock farming, and many communal uses including drinking water supply. In the present study, basic limnological parameters of ten irrigation reservoirs of Sri Lanka were investigated from June 2013 to February 2016, with a view to identifying influence of hydrological regimes on reservoir water quality. Spatio-temporal similarities of water quality parameters were studied employing the self-organizing map (SOM) routine of the artificial neural network application. The sample vectors, classified on the SOM lattice, indicated 6 clusters at 50% similarity level. When reservoir that were categorized according to hydraulic retention time (HRT) and relative reservoir level fluctuation (RRWL; defined as the ratio of the mean reservoir level amplitude to mean depth), were compared with dominant water quality parameters in SOM lattice, it was evident that some productivity-related water quality parameters were influenced HRT and RRWL. The results of the study revealed that HRT and RRWL can essentially be controlled through management of hydrological regimes in irrigation reservoirs, thus, close dialogue between irrigation authorities and other users of reservoir water resources are needed to ensure desired water quality of the reservoirs.