A bumblebee lands on a feeder laced with sugar water. The platform beneath it is uncomfortably hot. If the sugar is dilute, the bee leaves almost immediately. But if the sugar is concentrated enough, the bee stays, enduring the heat to keep drinking. That simple behavioral choice, repeated across hundreds of trials in a London laboratory, has become one of the most provocative data points in modern animal science: evidence that an insect with a brain smaller than a sesame seed may be weighing pain against reward, not just reacting on autopilot.
The finding is part of a series of controlled experiments, published between 2022 and 2025, that together build the strongest case yet that bumblebees experience something resembling pain. The work has survived peer review, a formal methodological challenge, and public reanalysis. As of June 2026, it is reshaping conversations about which animals deserve legal welfare protections and forcing a basic question back into the spotlight: how confident are we that insects feel nothing?
Three experiments, one converging picture
The motivational trade-off study, led by Matilda Gibbons and Lars Chittka at Queen Mary University of London, was published in Proceedings of the National Academy of Sciences in 2022. Researchers placed bumblebees (Bombus terrestris) on feeders set atop heated surfaces and varied both the temperature and the sweetness of the sucrose solution. Bees consistently tolerated more heat when the sugar reward was richer. A pure reflex would produce the same withdrawal every time the temperature crossed a threshold, regardless of what was on offer. Instead, the bees behaved as though they were making a cost-benefit calculation, adjusting their tolerance based on context.
A second experiment, published in iScience in 2024, looked at what bees do after being poked with a noxious stimulus on a single leg. Rather than launching into a generic escape or grooming themselves at random, the bees directed grooming specifically at the affected limb. That kind of body-directed protective behavior is difficult to explain as a simple alarm response. The raw behavioral data, including video coding schemes and exclusion criteria, were deposited on figshare, making the work unusually transparent and open to outside replication.
A third experiment, also published in iScience (doi:10.1016/j.isci.2024.111555), introduced a pharmacological test. Researchers gave bumblebees oral doses of gabapentin, a drug prescribed to humans for nerve pain that acts on central nervous system pathways in vertebrates. Bees that received the drug showed reduced heat sensitivity and weaker avoidance of noxious stimuli compared with controls. The result suggests that whatever neural machinery governs pain-like responses in bees may share functional overlap with vertebrate pain pathways, though the precise molecular targets in the bee brain have not yet been mapped.
Individually, each result is suggestive. Together, they satisfy several criteria that researchers use to distinguish nociception, the automatic detection of a harmful stimulus, from something closer to a felt experience. Context-dependent modulation, body-directed protection, and pharmacological sensitivity are all features that, in mammals, are associated with conscious pain processing. No single experiment proves subjective experience in a bee, but the convergence across three independent lines of evidence is what has made the scientific community pay attention.
The challenge and the rebuttal
Strong claims attract strong scrutiny, and the motivational trade-off finding got exactly that. Earlier in 2025, a formal methodological critique published in Animal Behaviour questioned whether the statistical interaction between heat and sucrose concentration truly supports a sentience-based interpretation. The critique’s authors are not named here because the DOI-linked paper itself provides full attribution; their argument was that alternative statistical models weakened the interaction effect and that simpler explanations, such as peripheral sensory adaptation or shifts in feeding motivation unrelated to pain, could account for the data.
Gibbons and colleagues responded with a reanalysis published alongside the critique in the same journal. Their reassessment confirmed the key interaction effect under their preferred modeling approach and addressed concerns about how trial order and individual differences among bees were handled. The exchange is a textbook example of healthy scientific friction: the original finding emerged stronger for having been tested, but the broader question of whether behavioral flexibility equals felt pain remains genuinely open. Accepting that bees show complex, adaptive behavior does not require committing to a particular theory of insect consciousness.
What the data still cannot tell us
Several significant gaps remain. No study has yet recorded direct electrophysiological or calcium-imaging data from bee nociceptors during the trade-off task itself. Without that neural-level confirmation, the behavioral evidence, however compelling, cannot definitively prove that bees feel pain in any way comparable to mammals.
Researchers also do not know whether the targeted grooming response persists or changes after repeated noxious stimuli over days. That kind of longitudinal data would help clarify whether the behavior is a short-term reaction or part of a longer-term protective strategy resembling sensitization or chronic pain.
There is also a question of generality. Nearly all of the detailed experiments have been conducted on Bombus terrestris under laboratory conditions. Whether similar motivational trade-offs and body-directed grooming occur in honeybees, solitary bees, or other insect orders is unknown. Comparative work would help determine whether pain-like states are a specialized feature of social bees or part of a broader pattern across insects. Until those data exist, sweeping claims about “insect sentience” as a category remain provisional.
The evidentiary framework that many researchers reference when evaluating invertebrate sentience was originally built for a different group of animals. A UK government-commissioned review led by Jonathan Birch at the London School of Economics laid out criteria for assessing pain-relevant sentience in cephalopod molluscs and decapod crustaceans. That analysis, which weighed evidence on learning, motivational trade-offs, and responses to analgesics, informed the inclusion of lobsters and octopuses in the UK’s Animal Welfare (Sentience) Act of 2022. No peer-reviewed paper has yet formally scored the bumblebee datasets against those same criteria in a structured assessment, leaving the policy bridge between the crustacean evidence and the insect evidence partly informal.
Why it matters beyond the lab
For anyone involved in agriculture, pest management, or animal welfare policy, the practical implications are not abstract. If insects can experience pain-like states, current regulations that exclude them from welfare protections may need revision. The UK has already expanded its sentience laws to cover crustaceans and cephalopods on the basis of structured evidence reviews. Similar formal assessments for insects would likely focus on bees first, given the existing experimental record.
In the meantime, a precautionary approach would favor minimizing unnecessary noxious procedures in research and industry, redesigning experiments to use the lowest effective stimulus intensities, and considering alternative methods where feasible. Lars Chittka, the senior researcher behind much of this work and author of The Mind of a Bee, has argued publicly that the evidence already warrants treating bees with greater care, even before every philosophical question about their inner lives is resolved.
Where the burden of proof now sits
The bumblebee studies do not close the book on insect sentience. But they shift where the burden of proof sits. The old default assumed insects were biological machines until proven otherwise. The emerging question is different: do we have enough certainty that bees feel nothing to justify treating them as though they cannot suffer? As behavioral, pharmacological, and eventually neural evidence accumulates through the rest of 2026 and beyond, the answer will shape not just scientific practice but the ethical boundaries we draw around some of the most numerous animals on Earth.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
