Jugular catheters were inserted in nine male rats under general isofluorane anesthesia and the catheters were connected to a commercially available computerized blood sampling device (Accusampler). Blood samples (150 µl) were collected every 4 h during the first 24 h after surgery and every 12 h during the following 72 h until 94 h after surgery, when the animals were killed. All fecal pellets were collected at blood sampling. Serum corticosterone and fecal concentrations of immunoreactive corticosterone metabolites and immunoglobulin A (IgA) were quantified by ELISAs. In blood, high corticosterone concentrations (.200 ng/ml) were recorded in the first samples obtained after surgery, but the concentrations decreased steadily during the day and became cyclical, showing a diurnal variation with high levels during evenings and low levels in the mornings. The automatic blood sampling itself did not result in recordable increases in serum corticosterone concentrations. The time delay between the presence of elevated corticosterone levels in blood and in feces was approximately 12 h. Fecal immunoreactive corticosterone metabolite levels remained elevated during the 94 h study period after surgery. The fecal concentrations of IgA showed substantial between-animal variation and decreased non-significantly after the surgery. Like serum corticosterone, fecal IgA showed a diurnal variation in amounts excreted, in this case with high values in the morning and low values in the evening. The concentrations of fecal corticosterone and IgA were negatively correlated in samples obtained before surgery but no correlation existed after surgery. This indicates that fecal immunoreactive corticosterone metabolites, but not IgA, constitute a good marker of acute stress. For immunoreactive corticosterone metabolites as well as for IgA, the concentration in feces correlated well with total excretion, making single fecal samplings usable as a measure of total secretion.