Recent studies have shown that cycling hypoxia may also be a factor in selecting and promoting cells with stem cell-like phenotype, presenting increased tumor-initiating capabilities and metastatic potential. Collectively, these data suggest that cycling hypoxia, within the tumor mass, may not only cause resistance to conventional therapies, but may also facilitate a more aggressive phenotype of tumor cells. These suggestions have been supported by the clinical observation that higher degree of tumor reoxygenation after radiotherapy is associated with worse patient prognosis. There are also reports showing that individual tumors differ depending on the extent of cycling hypoxia regions. Tumors that present interchangeable states of hypoxia and reoxygenation, present an increased metastatic potential and are more radioresistant. Identification of these tumors would inevitably improve cancer prognosis and enable treatment of patients with therapy tailored to each individual case. Thus, identifying the molecular pathways and genes involved in promoting the aggressive phenotype of tumor cells under cycling hypoxia conditions seems crucial. Herein, we investigated the influence of cycling and chronic hypoxia on gene expression profile in three cancer cell lines, using a microarray platform. The analysis indicated that cycling hypoxia exerts a similar, although weaker, influence on gene expression in cancer cells than chronic hypoxia. The main differences observed between the two types of hypoxia involved the expression of several genes such as IL-8, CXCL2, EPHA2, AREG, HBEGF, and PLAU, which are relevant to tumor progression. Our results indirectly suggest that cycling hypoxia may promote an aggressive phenotype by inducing the expression of genes regulating the immune response, invasion, and proliferation. There are two major types of oxygen kinetics detected in tumor mass, depending on time scales of oxygen fluctuations, the rapid oxygen kinetic and the slower oxygen kinetic. While the first one is relatively well-known, the latter is not precisely characterized. Here, we present a gene expression analysis in three cancer cell lines exposed to either chronic or cycling experimental hypoxia, or grown in control conditions. The cell lines used in this study represent three types of cancer, known to be affected by hypoxia. The cycling conditions were BEZ235 distributor chosen to mimic oxygen fluctuations in a slower time scale. In vivo, they occur during the remodeling of the vascular network when the hypoxic and reoxygenation periods.