The potential use of adenoviruses in therapy against cancer has evoked a rapidly moving field of research. Unlike conventional gene therapy vectors, oncolytic adenoviruses retain the ability to replicate. However, replication is restricted as much as possible to tumor cells, with the aim of eliminating these cells through viral cytotoxicity. The two key issues are to improve the efficiency of virus replication and cell killing while ensuring the specificity of these activities for tumor cells. Wild-type adenoviruses as such may already be usable for cancer therapy. Strategies to further improve efficiency and specificity include the partial or complete removal of viral genes. The idea is that functions carried out by the corresponding gene products are not required for replication in tumor cells, but are needed in normal cells. Accordingly, the removal of genes encoding E1B-55 kDa or E1B-19 kDa, or the mutation of E1A may improve the selective killing of tumor cells. On the other hand, the overexpression of the adenovirus death protein (ADP) may enhance viral spread and oncolytic efficiency. Other strategies to improve the specific oncolytic activity of replicating adenoviruses have been pursued. For instance, some promoters are active specifically in tumor cells, and these promoters were introduced into the viral genome, to regulate essential viral genes. Moreover, replicating viruses were engineered to express toxic proteins or drug converters. A number of these viruses have been tested successfully using tumor xenografts in nude mice as a model system. An oncolytic adenovirus lacking the E1B-55 kDa gene product, termed dl1520 or ONYX015, was injected into squamous cell carcinomas of head and neck in phase II clinical trials, and the results were encouraging when chemotherapy was applied in parallel. In the future, further progress might be achieved on the level of virus constructs, but also by refining and adjusting simultaneous conventional therapies, and by standardizing the assessment of the clinical outcome. Recent progress has been made towards the use of replicating virus constructs in cancer therapy. The goal of these developments is to remove cancerous cells from patients with the help of viruses that selectively replicate in these cells. These viruses are generally termed oncolytic viruses. Some convenient properties of adenovirus make this virus particularly useful for this purpose. It infects a large number of human cell types, especially epithelial cells, which give rise to the vast majority of human malignancies. It can be grown easily and to high titers, and the creation of virus recombinants is well established. Finally, a large body of basic research has already been carried out on this virus, facilitating its manipulation. Various approaches to use adenovirus as a cancer drug have been reviewed (Alemany et al. 1999a, 2000; Curiel 2000; Galanis et al. 2001b; Gromeier 2001; Heise and Kirn 2000; Kirn 2000a; Kirn et al. 2001; Kirn and McCormick 1996; Smith and Chiocca 2000; Sunamura 2000; Wells 2000; Wodarz 2001). The aim of this chapter is to provide an integrated overview of these strategies.