AbstractIn this study, an optimization methodology is followed in order to explore better form of heat sinks which improve thermal performances. Optimum designs of plate fin heat sinks (PFHSs) and modified shaped plate fin heat sinks (MS-PFHSs) are numerically investigated. The objective functions are minimizations of base plate temperature, entropy generation and mass. For both PFHSs and MS-PFHSs, optimization variables include inlet velocity (Vin), fin height (Hfin), and number of fins (NL). Plate fin form is adjusted for MS-PFHSs by adding two optimization variables: the rib height (Hrib) and the number of patterns in the flow direction (Wp). For the multi-objective optimization problems, the maximum base plate temperature limit (Tbase<70 °C) is used. The multi-objective genetic algorithm (MOGA) is used to solve optimization problems, and three-dimensional parametric models for numerical optimization work are examined using the finite volume approach. The flow is steady, incompressible, and turbulent, and heat transfer in the heat sink is represented by conjugate heat transfer (CHT). It is shown that MS-PFHSs outperform in terms of the analyzed objective functions. For the optimum designs, Tbase values of MS-PFHS and PFHS are 60.23 °C and 65.25 °C, respectively, while the mass values are same. The results also indicate that Tbase obtained in the optimum design of MS-PFHS is 7.69% lower than that obtained in the optimum design of PFHS.