This is a progress report on the development of an “ideal” in vivo dosimetry system for both experimental and routine clinical use in radiation therapy of cancer. During rigorous clinical studies this system has thus far proved to be unusually dependable, practical, versatile, and economical to operate. It has been applied in nearly every anatomical structure in living subjects. Further, it is promptly available for general application in clinical practice. The basic dosimeter consists of a tiny photoluminescent glass rod (1 × 6 mm.) encased in a miniature gold cartridge which we designed (1.6 × 10 mm.), with a removable screw-cap. Placed within or upon the body and exposed during radiation therapy, the glass rod may be removed at any time and “read” in a chamber which may be easily constructed or purchased from the commercial source which now supplies the glass rods. Unlike other dosimeters, “reading” does not discharge the unit and it may be replaced in the body for cumulative dose recording as high as 20,000 r (or as low as 20 r). This recording is permanent until erased by a simple heating procedure. We have employed these gold-shielded dosimeters successfully in nearly every body tissue and space in cancer patients receiving ionizing radiation from almost every kind of source, including 260 kv to 2,000-kv apparatus, cobalt 60, cesium 137, radium, radon, and iridium 192. In the Radiation Clinic, the gold-shielded glass rod reads accurately to within ±5 per cent; in the laboratory, within ± 2 per cent. Complete energy independence was noted in the range employed, so that this dosimeter does not have to be calibrated against each specific therapy source. For this advantage, and many others, the gold shield is indispensable. Without it, for example, there is striking energy dependence (and unreliability) below the supervoltage energy range; when portals are larger than 150 sq. cm.; and in tissue depth greater than 8 cm. When utilized in the bare state, the glass rods are not identifiable and are easily contaminated, chipped, and cracked, delivering false readings. We have positioned the gold-shielded microdosimeter in the heart, brain, paranasal sinuses, nasopharynx, pharynx, tongue, floor of mouth, lymph nodes, mandible, and skeleton. We placed the tiny chamber in the esophagus, larynx, trachea, bronchi, stomach, colon, rectum, renal pelvis, ureter, urethra, bladder, uterus, cervix, parametrium, prostate, mediastinum, aorta, vena cava, and in the general vascular system accessible to catheterization. For most applications, a common No. 8 polyvinyl feeding catheter has been found quite adequate, with dosimeters loaded singly or in tandem. For interstitial use, a large-bore needle was designed to carry the polyvinyl catheter or a hollow radium needle of appropriate length was employed.