Computational intelligence (CI) algorithms are designed to derive the biological behavior of various natural species. The evolving, behavioral, and adapting nature of the species are taken as the basis for the components of CI viz., Evolutionary Algorithms (EAs), Fuzzy Logic (FL), and Neural Networks (NN). The EAs, FL, and NN have their algorithmic structure, merits, and demerits. Hybridization is the most open research problem, as the hybridized versions of the algorithms could balance the merits and demerits of the constituent algorithms and show improved performance. The study presented in this paper is an attempt towards the hybridization of an EA called Genetic Algorithm (GA) and FL. Initially, various existing approaches to hybridizing GA and FL are explored. This led to the identification of balancing the population diversity of GA by a Fuzzy Logic Inference System (FLIS) as a research problem. This paper proposes a HyAGaFlis (Hybrid Algorithm of GA and FLIS) algorithm which adapts its crossover and mutation probabilities using a FLIS, to maintain the population diversity to reach the global optimum value without trapping in local optimum value(s). The performance of the classical GA and the HyAGaFlis algorithms are compared on a multimodal benchmarking function. The comparative study revealed that the proposed HyAGaFlis outperforms the classical GA in reaching the global optimum without loss of population diversity. The proposed algorithm is useful to practitioners and scientists who intend to use GA to solve their optimization problems, with less effort for GA’s parameter tuning. The novelty of the HyAGaFlis algorithm lies in the way how it controls the population diversity during the evolution process of GA.