About 15% of the ionizing radiation exposure to the general public comes from artificial sources, and almost all of this exposure is due to medical radiation, largely from diagnostic procedures. Of the approximately 3 mSv annual global per caput effective dose estimated for the year 2000, 2.4 mSv is from natural background and 0.4 mSv from diagnostic medical exams. Diagnostic and therapeutic radiation was used in patients as early as 1896. Since then, continual improvements in diagnostic imaging and radiotherapy as well as the aging of our population have led to greater use of medical radiation. Temporal trends indicate that worldwide population exposure from medical radiation is increasing. In the United States, there has been a steady rise in the use of diagnostic radiologic procedures, especially x rays. Radiotherapy also has increased so that today about 40% of cancer patients receive some treatment with radiation. Epidemiologic data on medically irradiated populations are an important complement to the atomic-bomb survivors' studies. Significant improvement in cancer treatment over the last few decades has resulted in longer survival and a growing number of radiation-related second cancers. Following high-dose radiotherapy for malignant diseases, elevated risks of a variety of radiation-related second cancers have been observed. Risks have been particularly high following treatment for childhood cancer. Radiation treatment for benign disease was relatively common from the 1940's to the 1960's. While these treatments generally were effective, some resulted in enhanced cancer risks. As more was learned about radiation-associated cancer risks and new treatments became available, the use of radiotherapy for benign disease has declined. At moderate doses, such as those used to treat benign diseases, radiation-related cancers occur in or near the radiation field. Cancers of the thyroid, salivary gland, central nervous system, skin, and breast as well as leukemia have been associated with radiotherapy for tinea capitis, enlarged tonsils or thymus gland, other benign conditions of the head and neck, or benign breast diseases. Because doses from diagnostic examinations typically are low, they are difficult to study using epidemiologic methods, unless multiple examinations are performed. An excess risk of breast cancer has been reported among women with tuberculosis who had multiple chest fluoroscopies as well as among scoliosis patients who had frequent diagnostic x rays during late childhood and adolescence. Dental and medical diagnostic x rays performed many years ago, when doses were presumed to be high, also have been linked to increased cancer risks. The carcinogenic effects of diagnostic and therapeutic radionuclides are less well characterized. High risks of liver cancer and leukemia have been demonstrated following thorotrast injections, and patients treated with radium appear to have an elevated risk of bone sarcomas and possibly cancers of the breast, liver, kidney, thyroid, and bladder.