The Bees Algorithm (BA) is a recent and powerful foraging algorithm which imitates the natural behaviour of bees. However, it suffers from certain limitations, essentially in the initialization step of the research areas, which is generally random and depends on the individuals' number in the population. In order to solve this problem, this paper proposes a novel hybrid optimisation approach, namely a Hybrid Firefly Bee Algorithm (HFBA), by using the Bees Algorithm (BA) and the Firefly Algorithm (FA). The FA is a swarm intelligence technique based upon the communication behaviour and the idealized flashing features of tropical fireflies. The proposed approach uses a FA in initialization step for a best exploration and detection of promising areas in research space. The performance of HFBA was investigated on a set of benchmark functions and compared with BA, and other well-knows methods. The results show that the HFBA has improved the computational time. It is also very efficient in finding optimal or near optimal solutions, and outperforms the other algorithms in terms of accuracy and speed.