The task of designing product forms is becoming increasingly challenging as consumers demand visually pleasing products that meet functional requirements. Researchers have been working on product concept generation systems that support designers in exploring design spaces to produce large numbers of product concepts from which possible solutions can be selected and developed. Such systems typically utilize the searching capabilities of metaheuristic techniques to explore solution spaces and generate variations of product forms. Recent research work in this domain has focused on hybrid approaches that combine metaheuristic techniques with other methods such as shape description and neural network approaches. This paper provides a review of the application, effectiveness and prospects for these approaches.