Although capacity has been extensively studied in wireless networks, most of the results are for homogeneous wireless networks where all nodes are assumed identical. In this paper, we investigate the capacity of heterogeneous wireless networks with general network settings. Specifically, we consider a dense network with n normal nodes and m = nb (0 < b < 1) more powerful helping nodes in a rectangular area with width b(n) and length 1/b(n), where b(n) = nw and -1/2 < w Â¿ 0. We assume there are n flows in the network. All the n normal nodes are sources while only randomly chosen nd (0 < d < 1) normal nodes are destinations. We further assume the n normal nodes are uniformly and independently distributed, while the m helping nodes are either regularly placed or uniformly and independently distributed, resulting in two different kinds of networks called Regular Heterogeneous Wireless Networks and Random Heterogeneous Wireless Networks, respectively. In this paper, we attempt to find out what a heterogeneous wireless network with general network settings can do by deriving a lower bound on the capacity. We also explore the conditions under which heterogeneous wireless networks can provide throughput higher than traditional homogeneous wireless networks.