Melatonin immunofluorescence in the rat colon was measured with a cadmium sulfide photometer in conjunction with an immunohistological double antibody procedure. The method proved to be very reliable with 5-10% intra- and interassay reliabilities. In addition, manipulation of the primary antibody concentration, the incubation time, and the nature of the first antibody produced predicted changes in fluorescence intensity. Treatment with p-chlorophenylalanine led to a marked reduction of melatonin-specific fluorescence.