Summary—We defined the SIMS conditions for radioiodine detection in animal and man thyroid follicles, in tissue sections (3 μm) chemically fixed and resin embedded. Two radioisotopes were tested:125I and129I, of high (14 mCi125I μg−1) and low specific activity (1.07 10−6mCi129I μg−1). In animal study, Wistar rats fed a normal iodine diet (10 μg127I day−1) were injected ip 24 h before sacrifice either with125I (7 10−3μg) or with129I at a dose identical to iodine diet (10 μg) or 3 times higher (30 μg). No SIMS signal of125I was obtainedin vivodue to its too low concentration, while radioiodine distribution was evidenced with both doses of129I. Local concentration of previously stored127I in follicular lumen was not modified, when compared to control (4.14 ± 0.03 μg/mg, m ± SE), by125I or129I at a dose of 10 μg, but was nearly doubled with129I at a dose of 30 μg, proof of a pharmacological effect on thyroid iodine regulation. In human study129I was excluded due to its long half‐life (1.6 107years), and125I was tested onlyin vitroon two surgical specimens of normal perinodular thyroid tissue maintained in mini‐organ culture for 48 h in presence of 100 μCi/ml of125I. The125I was detectable, its concentration was 1000‐fold higher than that of127I (1.5 ± 0.004 μg/mg). For bothin vivoandin vitrostudies, a positive correlation exists between newly organified radioiodine (125I or129I) and previously stored iodine (127I). In conclusion, the radioisotope129I used at a dose of 10 μg is well adapted for SIMS detectionin vivoof thyroid exchangeable iodine, but only in animal, while125I isotope is especiall