Abstract The characteristics of [[ 3H]neurotensin binding were studied using membranes prepared from the rat brain. Binding of [ 3H]neurotensin was found to be specific, saturable and reversible. Under the conditions of the assay non-specific binding represented less than 20% of the total binding at a radioligand concentration of 2 nM. The specific [ 3H]neurotensin binding increased linearly with protein concentration and was dependent on the pH and the temperature of the incubation medium. At 25 °C equilibrium was reached rapidly and the association kinetics appeared to be monophasic. The dissociation was not monophasic and it could be resolved into two distinct components. Scatchard analysis of the saturation data indicated a single population of binding sites with a density of 432 fmol/mg protein and an equilibrium dissociation constant of 2.85 nM. A Hill transformation of the competitive inhibition of specific [ 3H]neurotensin binding by increasing concentrations of unlabelled peptide yielded a slope not significantly different from unity. Neurotensin 1–13 and various neurotensin analogues were tested for their ability to compete with [ 3H]neurotensin for its binding site. Neurotensin 1–13, neurotensin 8–13 and the amphibian skin peptide xenopsin were equipotent and strongly inhibited the specific binding of [ 3H]neurotensin. Neurotensin 9–13 and the chicken intestinal peptide Lys 8, Asn 9-neurotensin 8–13 were weakly active, whereas neurotensin 10–13 and the amino-terminal fragments neurotensin 1–6, neurotensin 1–8 and neurotensin 1–11 were inactive. Physiological concentrations of sodium choloride inhibited specific [ 3H]neurotensin binding, whereas divalent cations and guanyl nucleotides did not produce significant change in either the equilibrium dissociation constant or the total number of binding sites. There was no apparent correlation between the content of neurotensin-like immunoreactivity in the rat brain and the density of [ 3H]neurotensin binding sites in the various brain regions. The highest density of binding sites was found in the hypothalamus and the frontal cortex, intermediate levels in striatum, thalamus, midbrain, hippocampus and olfactory bulb, and low levels in cerebellum and pons-medulla oblongata. In general, there was less variation between different brain regions in the number of [ 3H]neurotensin binding sites than in the content of neurotensin-like immunoreactivity. The characteristics of this binding assay are consistent with [ 3H]neurotensin binding to a physiological receptor.