The traveling salesman problem is one of the most well-solved hard combinatorial optimization problems. Any new algorithm or heuristic for the traveling salesman problem is empirically evaluated based on its performance on standard test instances, as well as on randomly generated instances. However, properties of randomly generated traveling salesman instances have not been reported in the literature. In this paper, we report the results from an empirical investigation on the properties of randomly generated Euclidean traveling salesman problem. Our experiments focus on the properties of the edge lengths and the distribution of the tour lengths of all tours in instances for symmetric traveling salesman problems.