Algal bloom has been a persistent problem for both fresh and marine water, with no exception for a coastal reservoir (CR). Among the algal bloom mitigations for a CR, shape optimisation to reduce algal bloom occurrence has been frequently mentioned. However, there was no literature found on the actual shape optimisation study or process for CR. Thus, this research was done to bridge this gap, particularly for the second-generation CR. Hydrodynamic model of MIKE 21 has been used, with secondary data obtained from published papers and Google Earth. The secondary data of critical velocity corresponding to algal growth were only available for cyanobacteria, chlorella, filamentous algae and phytoplankton. Hence, only these algae species were considered in the algal mapping. All models were simulated for eight idealised cases of second-generation CR at the Yangtze Estuary. These different geometric shapes were analysed and compared, considering several factors including the average velocity in the reservoir, presence of stagnant water, percentage of occurrence for each algae species and so on. From the results, the reservoir model with the shape of “shorter piano key” ranked the highest among all the shapes in terms of proneness to algal bloom, based on the flow velocity within the reservoir. From the findings, further shape optimisation was done on second-generation CRs. From shape optimisation process, the optimised shape of the “little dinosaur” and “little pencil” showed excellent reduction in algal blooming. However, “little dinosaur” was preferred as its location for algal bloom is small and controllable due to the presence of “piano key” structure. Lastly, all the findings were applied to an existing CR at Qingcaosha to see if shape optimisation based on the analysis can reduce areas prone to algal blooming. The optimised model of Qingcaosha showed great reduction on area prone to algal blooming compared to its original shape but the addition of “piano key” structure did not have significant impact on the reduction of algal bloom occurrence since the shape of Qingcaosha is highly dependent on its natural topography. From the study, it was concluded that shape optimisation for topographic-dependent CR should be done on case-by-case basis, following the flow direction in the reservoir. As for second-generation CR without topographical constraints, the shape optimisation can be done by placing inlet perpendicular to the flow direction, minimising corners, implementing piano-key-like structure, optimising shape based on flow direction and refer “little dinosaur” or “little pencil” for the overall shape optimisation design.