A reliable method for measuring lymph flow in physiological units would be valuable, especially in conditions in which it is uncertain whether lymph flow is increased or decreased. The requirements of a radiopharmaceutical for such measurement include stable radionuclide labelling and rapid access to lymphatic vessels following tissue injection but no access to blood vessels. A soluble macromolecule is likely to come closest to meeting these requirements. Technetium-99m-labelled human polyclonal immunoglobulin (HIG) was therefore investigated firstly in comparison with 99mTc-labelled human serum albumin (HSA) in patients undergoing routine lymphoscintigraphy and secondly with respect to injection site in a group of volunteers with post-mastectomy oedema (PMO). Subcutaneous injection of 99mTc-HIG into the web space of a distal extremity gave images in which lymphatic vessels were more clearly defined compared with images obtained after injection of 99mTc-HSA. Lymph nodes were also more clearly defined, suggesting specific retention of HIG, possibly through Fc-mediated binding. Peripheral blood sampling showed a delayed arrival in blood of radioactivity after 99mTc-HIG compared with 99mTc-HSA, although ultimately, the blood recovery of 99mTc-HIG was significantly higher (P < 0.05) than that of 99mTc-HSA. Clearance rates of radioactivity from the injection site were not significantly different, however, between the two agents. In patients with PMO, web space injection of 99mTc-HIG gave excellent images of normal lymphatic vessels, of lymph nodes and of abnormal lymph drainage such as dermal backflow in swollen arms. In contrast, neither lymphatic vessels nor lymph nodes were visualised after injection into the skin of the dorsum of the distal forearm. Although there was no difference in clearance rates from the injection sites between normal and swollen arms with either agent in PMO, clearance was significantly faster following injection into the web space (0.11% per minute for normal and swollen arms combined) than into the forearm (0.053% per minute; P < 0.05). These results suggest that (a) 99mTc-HIG is a potentially useful agent for measuring lymph flow and lymph node function; but (b) injection into the dorsum of the forearm is not a useful method of administration for these measurements; and (c) clearance rates from the injection site do not support the notion that PMO is the result of decreased lymph flow. Further studies are warranted to evaluate 99mTc-HIG as an agent for assessment of lymphatic function, especially with respect to measurement of lymph flow and possibly also for the evaluation of lymph node Fc-mediated immunocompetence.