Cuckoo search (CS) belongs to a class of novel nature-inspired algorithm inspired by the obligate brood parasitism of some cuckoo species by laying their eggs in the nests of other host birds (of other species). CS has been successfully applied to solve continuous optimization problems, however, its potential in discrete problems has not been sufficiently explored. Basing on the present algorithm of CS, this paper construct an discrete cuckoo search (DCS) to solve traveling salesman problem (TSP). Two schemes are studied, i.e. discrete step size and the cuckoo's updating scheme. DCS is evaluated on the TSP problem, comparing performance to other methods. Experimental results shows that the proposed algorithm performs very well to solve some simple TSP instances, but it can be trapped into local optimum solutions for some other instances.