Fish meet most scientific criteria proposed for sentience, including the ability to detect and respond to painful stimuli, form memories of positive and negative experiences, and display flexible behaviors that go well beyond simple reflexes. The evidence has accumulated across dozens of species and multiple branches of research, from electrophysiology and stress hormone measurement to studies of social behavior and self-recognition. What makes the question interesting is not whether any single study proves the case, but how many independent lines of evidence point in the same direction.
Fish Have the Hardware for Pain
One of the foundational questions in animal sentience is whether an organism can detect harmful stimuli at all. In 2003, researchers at the University of Edinburgh recorded electrical signals from the trigeminal nerves of rainbow trout and found specialized receptors on the head that responded to mechanical pressure, dangerously high temperatures (above 40°C), and acetic acid, a known irritant. These receptors, called polymodal nociceptors, behave much like the pain-sensing nerve fibers found in mammals, though with a notable anatomical difference: trout have more A-delta fibers (the fast-conducting type) than C fibers (the slow type that dominates in mammals).1PubMed Central. Do fishes have nociceptors? Evidence for the evolution of a vertebrate sensory system That structural difference is informative about how fish nociception evolved, but the core finding is straightforward: the sensory equipment for detecting tissue damage exists in fish brains.
The brain regions that process this information also share more with mammalian brains than was once assumed. The fish pallium, the outer region of the forebrain, contains areas that appear to serve functions parallel to the mammalian hippocampus and amygdala. Experimental work has shown that the medial pallium is involved in avoidance learning and the lateral pallium handles spatial learning, mirroring the division of labor seen in land vertebrates.2PubMed Central. Spatial Cognition in Teleost Fish: Strategies and Mechanisms Opioid receptors and the body’s own painkillers (endogenous opioids) are present in fish brains and spinal cords, and morphine blocks avoidance learning in fish in the same way it does in mammals. The electrophysiological properties of fish nociceptors are nearly identical to those of higher vertebrates, suggesting these features evolved before fish and land animals diverged hundreds of millions of years ago.
Perhaps most directly, researchers have shown that consciousness, or its loss, can be measured in fish using electroencephalogram recordings. When rainbow trout were exposed to an anesthetic, non-invasive EEG showed them transitioning to surgical depth of anesthesia within five minutes, with measurable changes in brain signal amplitude and the disappearance of visually evoked responses.3Aquaculture Research. Non‐invasive recording of brain function in rainbow trout: Evaluations of the effects of MS‐222 anaesthesia induction The fact that fish brains show graded states of awareness, responsive to the same classes of drugs that alter consciousness in mammals, is hard to square with the idea that fish are merely running on autopilot.
Behavioral Responses That Go Beyond Reflex
If fish only had the nerve fibers to detect damage, skeptics could argue that this is mere nociception without subjective experience. But the behavioral evidence makes a simple-reflex interpretation difficult to sustain. When trout are injected with a mild acid solution on the lip, they show prolonged changes in behavior: rocking back and forth, rubbing the affected area against the tank wall, and losing interest in food. These responses last far longer than a reflexive withdrawal would, and they get worse in proportion to the severity of the stimulus. Fish treated with a painful stimulus also perform competing tasks less well, as if their attention has been redirected.4PubMed Central. Evolution of nociception and pain: evidence from fish models
Goldfish show something even more telling: they weigh the costs and benefits of pain avoidance against other motivations. In experiments where food was placed in a zone that also delivered an electric shock, goldfish reduced their feeding attempts as shock intensity increased. But when they were hungrier, they tolerated more shock to get the food, and their escape responses became more pronounced as they entered the danger zone, suggesting they were aware of the risk and adjusting their behavior accordingly.5Applied Animal Behaviour Science. Trade-offs between feeding and shock avoidance in goldfish (Carassius auratus) A simple reflex cannot weigh hunger against pain. That kind of flexible decision-making implies central processing of the unpleasant nature of the experience.
Further evidence comes from studies in which fish altered their learned avoidance behavior depending on the social context. Goldfish and trout that had learned to avoid a painful stimulus changed their avoidance response when a companion fish was present, suggesting that even a supposedly automatic response to pain is modulated by social factors rather than running on a fixed script.6Applied Animal Behaviour Science. Avoidance learning in goldfish (Carassius auratus) and trout (Oncorhynchus mykiss) and implications for pain perception
Analgesics Work on Fish, Which Tells Us Something
If fish were simply responding to tissue damage with mechanical reflexes, painkillers should have no effect on their behavior. But morphine does change how fish behave after a noxious event. In zebrafish exposed to acetic acid, those pre-treated with morphine showed significantly higher activity levels than untreated fish, meaning the morphine reduced the behavioral suppression caused by the painful stimulus.7PubMed Central. A Novel Behavioral Fish Model of Nociception for Testing Analgesics The effect was dose-dependent, becoming more pronounced in the second hour of observation.
That said, the analgesic evidence is not as neat and tidy as researchers would like. A broad review of fish analgesia studies concluded that there is no overwhelming evidence of efficacy for most painkillers in fish, though morphine stands out as having the most consistent beneficial effects on both behavior and physiological stress markers.8PubMed Central. Updated Review of Fish Analgesia Part of the difficulty is methodological: it is harder to measure subjective pain relief in a fish than in a mammal that can be trained to press a lever. But the fact that opioids modulate fish responses to noxious stimuli at all is consistent with a system that is doing more than routing signals through a spinal reflex arc.
Stress Hormones Tell a Parallel Story
Fish, like mammals, release cortisol when they are stressed. Cortisol is the main glucocorticoid involved in the stress response in most fish species, and it has been used extensively to study the effects of everything from handling and transport to exposure to pollutants.9PubMed Central. Cortisol as a Stress Indicator in Fish: Sampling Methods, Analytical Techniques, and Organic Pollutant Exposure Assessments When fish experience something unpleasant, cortisol spikes. When they are in the presence of companions, cortisol levels can be lower, as described in the social buffering research below.
The complication is that cortisol alone does not prove conscious suffering. In one study of Nile tilapia and common carp, the “classical” stress markers like cortisol, glucose, and lactate levels could not distinguish between the pain of a tail-fin clipping procedure and the stress of being handled.10PubMed. Tailfin clipping, a painful procedure: Studies on Nile tilapia and common carp This does not mean fish do not feel pain; it means that stress hormones are blunt instruments that respond to many kinds of disturbance. Researchers increasingly rely on behavioral indicators alongside physiological ones to build a more complete picture.
Emotional Memory and Place Preferences
Sentience implies not just detecting a bad thing in the moment but remembering that it was bad and acting on that memory later. Fish demonstrate exactly this. In conditioned place preference experiments, fish that received a pleasant stimulus in one part of a tank later spent more time and moved more in that area, even when the stimulus was gone. Fish exposed to an unpleasant stimulus in a particular location avoided it afterward, spent more time in the opposite side, and showed elevated cortisol levels.11Applied Animal Behaviour Science. Use of conditioned place preference/avoidance tests to assess affective states in fish The fish retained memories of events with positive or negative emotional significance and retrieved those memories when environmental cues were present. This is the kind of evidence researchers use to assess affective states in mammals, and fish pass the same tests.
There is even suggestive evidence that fish engage in play. A study surveyed 66 species of fish for their responses to a laser pointer stimulus. A large majority, about 88%, showed attentiveness to the laser, with roughly 65% displaying moderate or high engagement. The most common behaviors included tracking the dot, chasing it, and attempting to catch it in their mouths.12PubMed Central. Gills Just Want to Have Fun: Can Fish Play Games, Just like Us? Play behavior in animals is widely considered a marker of positive emotional states, and while the laser study does not prove fish are having fun, the breadth of species that respond this way is notable. It is harder to explain as a reflexive feeding strike when so many species across different ecological niches do it.
Self-Recognition, Tools, and Cooperative Hunting
Some of the most striking evidence for fish cognition comes from tests traditionally associated with “higher” animals. The mirror self-recognition test, long considered a benchmark for self-awareness, has now been passed by cleaner wrasse. In these experiments, a small colored mark resembling a parasite was placed on the fish’s throat. Fish did not scrape their throats when the mirror was covered, and they did not scrape when a transparent (sham) mark was used. But when the mirror was exposed and a visible brown mark was present, every single fish in one set of experiments attempted to scrape it off.13PubMed Central. Further evidence for the capacity of mirror self-recognition in cleaner fish and the significance of ecologically relevant marks Follow-up work found that some cleaner fish began mark-directed behavior within 30 minutes of first seeing a mirror, faster than many other species known to pass the test.14PubMed. Rapid self-recognition ability in the cleaner fish Across all studies to date, cleaner wrasse show the highest pass rate of any non-human species tested, at about 94%.
Tool use, another behavior once thought unique to primates and a few birds, has been documented in multiple wrasse species. The sixbar wrasse has been observed repeatedly carrying food pellets too large to swallow to a specific rock and smashing them into smaller pieces, returning to the same “anvil” each time with consistent success.15PubMed. Tool-like behavior in the sixbar wrasse, Thalassoma hardwicke (Bennett, 1830) The graphic tuskfish in New Caledonia does the same thing with bivalves, sometimes striking a prey item against a rock or coral head as many as 28 times in a single session.16PubMed. Tool use by the graphic tuskfish Choerodon graphicus These observations suggest tool use may be widespread among wrasses, which is a large and diverse family of fish found in oceans around the world.
Cooperative hunting adds another dimension. In the Red Sea, groupers and giant moray eels hunt together in a remarkably coordinated way. Groupers actively signal to morays to initiate joint searches, and they recruit morays to specific hiding places where prey is concealed. The signaling depends on the grouper’s hunger level, and both species benefit from the arrangement.17PLoS Biology. Interspecific Communicative and Coordinated Hunting between Groupers and Giant Moray Eels in the Red Sea This kind of interspecies communication and intentional coordination is rare in the animal kingdom and challenges the assumption that fish behavior is primarily instinctive.
Social and Emotional Lives
Fish are often thought of as solitary or, at best, as schooling animals driven purely by safety-in-numbers instincts. The reality is more complex. Social buffering, where the presence of companions reduces an individual’s stress response, has been demonstrated in multiple fish species. Three-spined sticklebacks exposed to a simulated predator had higher cortisol levels when alone than when in a group. Zebrafish showed the same pattern: those facing a novel, stressful environment alone froze more and had higher cortisol than those who had company.18PubMed Central. Social buffering of the stress response: insights from fishes The buffering effect even shows sex differences. In zebrafish, males decreased their freezing behavior more than females in the presence of calm companions, while males also responded more intensely than females when companions were distressed.19PubMed Central. Sex differences in social buffering and social contagion of alarm responses in zebrafish
Perhaps the most provocative social finding involves emotional contagion. Emotional contagion, where one individual’s distress triggers a matching emotional response in an observer, is considered the most basic form of empathy. Researchers studying zebrafish found that observer fish imitated the distressed behavior of companion fish, and this response was regulated by oxytocin, the same hormone that governs empathic behavior in mammals. Zebrafish mutants lacking functional oxytocin or oxytocin receptors did not show the contagion effect, while adding oxytocin back restored it. The brain regions involved in this process in zebrafish are homologous to those involved in emotional contagion in rodents.20PubMed. Evolutionarily conserved role of oxytocin in social fear contagion in zebrafish The implication is that the neural machinery for basic empathy is ancient, predating the split between fish and land vertebrates.
The Neocortex Objection
The main scientific argument against fish sentience centers on brain anatomy. In humans, the neocortex is essential for conscious experience, including the conscious experience of pain. Fish do not have a neocortex. Some researchers have argued that without this structure, fish simply cannot have subjective experiences, no matter how complex their behavior appears.21Animal Sentience. Lack of neocortex does not imply fish cannot feel pain
This argument has largely fallen out of favor among researchers who study fish cognition, for a few reasons. First, the fish pallium contains regions that are functionally analogous to mammalian structures involved in emotion and learning, even though they look different anatomically. Second, birds also lack a neocortex yet show sophisticated cognition and are widely accepted as sentient. Third, and most broadly, complex cognition has arisen multiple times independently across the vertebrate tree. If sentience tracks cognition at all, demanding one specific brain structure as a prerequisite ignores how evolution actually works.22Animal Sentience. Convergent evolution of sentience? Insisting on a neocortex as the ticket to consciousness is a bit like insisting that only animals with wings can fly, while ignoring bats. Different evolutionary lineages solve the same problem with different hardware.
What This Means for How Fish Are Treated
The practical stakes of the sentience question are enormous. Billions of fish are slaughtered in aquaculture every year, and the vast majority are not stunned before killing, meaning they are exposed to what the evidence suggests is considerable pain and distress.23PubMed Central. Stunning methods in aquaculture slaughter and their implications for fish welfare Even when stunning methods are used, there is limited understanding of how reliable those methods are or how to evaluate their welfare outcomes. The gap between what the science suggests about fish sentience and how the industry treats fish is wider than for almost any other farmed animal.
In research settings, a parallel lag exists. Fish are among the most commonly used animals in laboratory science, yet the welfare guidelines for their care were largely derived from standards developed for mammals. The result is a regulatory framework that may not be well suited to the specific needs and sensitivities of fish.24PubMed. Fish research and the institutional animal care and use committee As the evidence for fish sentience has grown, there has been increasing pressure to update these standards, but progress is slow and uneven across jurisdictions.
Spatial Navigation and Map-Like Memory
One dimension of fish cognition that does not always make it into the sentience discussion is spatial intelligence. Fish do not simply swim around randomly until they bump into food or a predator. Many species use sophisticated navigational strategies, including allocentric navigation, which means orienting based on an internal map of the environment rather than simply following a compass direction or body-centered cues. The dorsolateral pallium, the brain region considered analogous to the mammalian hippocampus, appears crucial for this ability. Fish with intact dorsolateral pallia can perform shortcuts and detours, implying they hold relational representations of their surroundings rather than just memorizing specific routes.2PubMed Central. Spatial Cognition in Teleost Fish: Strategies and Mechanisms This capacity matters for the sentience discussion because flexible, map-based navigation is the kind of cognitive ability that is difficult to explain without some form of internal experience or at least a rich representational inner life. A creature that can mentally rotate its environment, calculate a novel shortcut, and remember the functional qualities of specific objects within that environment is doing something qualitatively different from running a set of stimulus-response rules.