The recent improvements of therapeutic approaches in oncology have allowed a certain number of patients with advanced disease to survive much longer than in the past. So, the number of cases with brain metastases and metastatic spinal cord compression has increased, as has the possibility of developing a recurrence in areas of the central nervous system already treated with radiotherapy. Clinicians are reluctant to perform re-irradiation of the brain, because of the risk of severe side effects. The tolerance dose for the brain to a single course of radiotherapy is 50-60 Gy in 2 Gy daily fractions. New metastases appear in 22-73% of the cases after whole brain radiotherapy, but the percentage of reirradiated patients is 3-10%. An accurate selection must be made before giving an indication to re-irradiation. Patients with Karnofsky performance status > 70, age < 65 years, controlled primary and no extracranial metastases are those with the best prognosis. The absence of extracranial disease was the most significant factor in conditioning survival, and maximum tumor diameter was the only variable associated with an increased risk of unacceptable acute and/or chronic neurotoxicity. Re-treatment of brain metastases can be done with whole brain radiotherapy, stereotactic radiosurgery or fractionated stereotactic radiotherapy. Most patients had no relevant radiation-induced toxicity after a second course of whole brain radiotherapy or stereotactic radiosurgery. There are few data on fractionated stereotactic radiotherapy in the re-irradiation of brain metastases. In general, the incidence of an "in-field" recurrence of spinal metastasis varies from 2.5-11% of cases and can occur 2-40 months after the first radiotherapy cycle. Radiation-induced myelopathy can occur months or years (6 months-7 years) after radiotherapy, and the pathogenesis remains obscure. Higher radiotherapy doses, larger doses per fraction, and previous exposure to radiation could be associated with a higher probability of developing radiation-induced myelopathy. Experimental data indicate that also the total dose of the first and second radiotherapy, interval to re-treatment, length of the irradiated spinal cord, and age of the treated animals influence the risk of radiation-induced myelopathy. An alpha/beta ratio of 1.9-3 Gy could be generally the reference value for fractionated radiotherapy. However, when fraction sizes are up to 5 Gy, the linear-quadratic equation become a less valid model. The early diagnosis of relapse is crucial in conditioning response to re-treatment.