Opioid receptors can be labeled with [3H]ligands in lightly fixed tissue sections mounted on microscope slides. The preparation of these sections does not seem to alter any of the known characteristics of opioid receptors. The light microscopic autoradiographic distribution of these binding sites can be observed if one attaches emulsion-coated coverslips to these slides to obtain autoradiograms. The distribution of [3H]diprenorphine binding sites determined by this in vitro method is identical to the distribution found in earlier studies utilizing in vivo labeling of opioid receptors. In addition, [3H]opioid peptide binding sites and [3H]dihydromorphine binding sites may be similar, perhaps identical, to those for [3H]diprenorphine, an opiate antagonist. This method has several important advantages over earlier methods for determining the autoradiographic localization of receptors. It is possible, by washing, to reduce nonspecific binding to low levels. Since receptor labeling is performed with single tissue sections mounted on slides, one can examine different receptors in different but adjacent sections. One can use ligands that are not entirely suitable for in vivo labeling (For example, one can use [3H]peptides which do not normally cross the blood-brain barrier). One can perform autoradiographic studies after labeling tissues under a wide variety of conditions (For example, one can examine the effects of various ions and nucleotides on ligand binding distributions). This technique seems to be free of various artifacts, such as edge artifacts, which were common by techniques used in our laboratory earlier. Since one does not have to load an entire animal with radioactive ligand as one must with in vivo labeling, this approach is economically advantageous. One can use postmortem tissues, including that from humans to study receptor distribution with a high anatomical resolution.