In fish and other vertebrate retinas, although dopamine release is regulated by both light and an endogenous circadian (24-hour) clock, light increases dopamine release to a greater extent than the clock. The clock increases dopamine release during the subjective day so that D(2)-like receptors are activated. It is not known, however, whether the retinal clock also activates D(1) receptors, which display a much lower sensitivity to dopamine in intact tissue. Because activation of the D(1) receptors on fish cone horizontal (H1) cells uncouples the gap junctions between the cells, we studied whether the clock regulates the extent of biocytin tracer coupling in the goldfish retina. Tracer coupling between H1 cells was extensive under dark-adapted conditions (low scotopic range) and similar in the subjective day, subjective night, day, and night. An average of approximately 180 cells were coupled in each dark-adapted condition. However, bright light stimulation or application of the D(1) agonist SKF38393 (10 microM) dramatically reduced H1 cell coupling. The D(2) agonist quinpirole (1 microM) or application of the D(1) antagonist SCH23390 (10 microM) and/or the D(2) antagonist spiperone (10 microM) had no effect on H1 cell coupling in dark-adapted retinas. These observations demonstrate that H1 cell gap junctional coupling and thus D(1) receptor activity are not affected by endogenous dopamine under dark-adapted conditions. The results suggest that two different dopamine systems are present in the goldfish retina. One system is controlled by an endogenous clock that activates low threshold D(2)-like receptors in the day, whereas the second system is controlled by light and involves activation of higher threshold D(1) receptors. J. Comp. Neurol. 467:243-253, 2003. Copyright 2003 Wiley-Liss, Inc.