Bat algorithm is one of the younger members of swarm intelligence algorithms that was introduced by Yang in 2010. So far, several variants based on this algorithm have been proposed for coping with the continuous and discrete problems. This paper introduces a novel self-adaptive bat algorithm (SABA) that borrows the self-adapting mechanism from self-adapted differential evolution known under the name jDE. This algorithm was tested on ten benchmark functions taken from publications. The obtained results were very promising and showed that this algorithm might be very suitable for use, especially, in the continuous optimization.