About 65 years ago, the first intracellular recordings ever to be made in the retina were made in horizontal cells (HCs) of goldfish (Carassius auratus). These recordings were some of the first examples of graded potentials (where membrane potential changes in proportion with a stimulus), and it was assumed that outer retinal neurons did not present action potentials (APs). HCs are interneurons which provide inhibitory feedback to photoreceptors, and in vision they are responsible for improving contrast, edge detection, and color opponency. Since these first recordings, Ca2+‐based APs and spontaneous increases in intracellular Ca2+ concentration ([Ca2+]i) have been found in fish HCs, although they are poorly understood and their role in vision (if any) is unknown. We characterized spontaneous, Ca2+‐based APs in isolated goldfish HCs by measuring changes in [Ca2+]i and membrane potential (Vm). Using Fura‐2 Ca2+ imaging, we found spontaneous Ca2+ activity characterized by transient elevation in [Ca2+]i of short (seconds) and long (minutes) duration, in 157/177 cells (89%). These APs were similar in amplitude to Ca2+ responses evoked by glutamate. APs were reversibly eliminated in Ca2+‐free solutions (n=18), and were abolished by the L‐type Ca2+ channel blocker, nifedipine (100 μM; n=5). APs were also dependent on intracellular Ca2+ stores: they were abolished by the ryanodine receptor antagonists, ryanodine (20 μM; in 7/8 cells) and dantrolene (50 μM; n=7). The ryanodine receptor agonist, caffeine (n=10), increased AP frequency 2.8‐fold (p<0.0002) and reduced median duration (29.3 s to 18.7 s), time to peak (12.3 s to 7.8 s), and area under the curve (by 59.6%). When caffeine was co‐applied with the store‐operated channel antagonist, 2‐aminoethyldiphenyl borate (2‐APB), frequency was unaffected, further confirming a role for stores. We tracked changes in Vm with whole‐cell current‐clamp recording, and with the voltage‐sensitive dye, FluoVolt. Phenotypes of APs in FluoVolt (n=6) and current‐clamp (48/57 cells; 84%) experiments displayed a sharp rise to peak, followed by a slow decline for the duration of the transient, and a steep return to baseline. In current‐clamp experiments, APs were dependent on L‐type Ca2+ channel activity. APs were blocked by Co2+ (5 mM; n=5) and nifedipine (100 μM; n=5), and were amplified and prolonged by the L‐type channel‐permeant ion, Ba2+ (15 mM; n=6). In addition, APs were abolished in 6/9 cells in Ca2+‐free solution. Collectively, our data suggest that activation of voltage‐dependent L‐type Ca2+ channels leads to Ca2+ influx, depolarization of Vm, and additional Ca2+ release from stores via ryanodine receptors. Understanding this phenomenon is a first step to elucidating a possible role for APs in vision or retinal physiology. This work also challenges the long‐held belief that outer retinal neurons do not present APs, and may lead to new paradigms about how visual information is encoded in the vertebrate retina.Support or Funding InformationWe acknowledge the support of the Natural Sciences and Engineering Research Council of Canada (NSERC), grant 342303.This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.