Large amounts of iodide from the body iodide pool are secreted with saliva, gastric juice and bile, and probably in pancreatic juice. The concentration of iodide in these secretions is proportional to the plasma iodide level. Secretion-iodide to plasma-iodide ratios vary greatly between individuals and between secretions in the same person. Iodide released from iodinated proteins broken down in the gut lumen can be reabsorbed and enter the body iodide pool even if regular doses of ion exchange resin are given. In these circumstances, no conclusions can be drawn concerning the source of iodide excreted with resin in the faeces. The resin test therefore cannot provide quantitative information about protein loss into the gut, with the possible exception of excessive leakage into the ileum or colon. Specific immunological assay for plasma proteins can be used to detect protein leakage into the gut. This technique has been used to demonstrate significant faecal losses of serum albumin in one patient. In a previous paper 1 we have presented evidence firstly, that ion exchange resins bind and retain iodide in the gut lumen, and secondly, that there is a relatively rapid passage of resin from stomach to colon. The present paper describes studies undertaken to check whether this rapid transit of resin influences the interpretation of results from experiments where resin is used to measure exogenous, or intra-gut catabolism of an intravenously administered radio-iodinated serum protein. Sites of iodide secretion into the gut lumen were investigated in normal people, patients and in dogs. Depending on the information required, radio-iodine, as free iodide, or bound to protein was given orally or intravenously. The radio-iodide content of the various gastrointestinal secretions was determined and, where possible, related to the plasma concentration of radio-iodide. The quantity of a radio-iodide dose excreted per day in the various secretions was assessed and the sum of these quantities was compared with the amount of radio-iodide bound to resin in stools of patients and normal subjects given intravenous radio-iodide. The probability of two-way movement of iodide across the gut and duodenal mucosae was investigated in dogs. Checks for the presence of four specific plasma proteins in gastrointestinal fluids were made by an immunological method, and in one patient given radio-iodinated 131I-human serum albumin (HSA) intravenously, the immunologically determined faecal excretion of albumin was compared with the resin radio-iodide excretion values.