To what extent do chickens suffer when gassed with CO2?

On 9 th June 2025, in an apparent world first, undercover footage of chickens being gassed to death, in the UK in 2024, was publicly released (Carbstrong, 2025). The footage depicts a CO2 gas pit system within which crated chickens are mechanically lowered into increasing concentrations of CO2 (Royal Society for the Prevention of Cruelty to Animals [RSPCA], 2025; p. 82). Specifically, the footage captured the CO2 stunning of 'spent' laying hens (Carbstrong, 2025)-laying hens of roughly 1.5 years of age who are no longer sufficiently productive to be of laying value to the egg industry. According to Carbstrong (2025), this includes hens from farms certified with the UK higher welfare label of RSPCA Assured. The Independent covered this news story and featured commentary from an RSPCA Assured spokesperson who had seen a longer nine-minute clip (Dalton & Colley, 2025). According to The Independent, the RSPCA spokesperson dismissed any welfare concerns arising from the footage, stating: "the birds in the footage were already unconscious and were not in pain". The following short article seeks to highlight the flaws within this blanket statement, and to suggest what the RSPCA could instead be advocating, as a leading UK animal welfare organisation and a provider of higher welfare labelling (RSPCA Assured, 2025a). The onset of loss of consciousness in chickens during CO2 gassing CO2 gas stunning/slaughter has become the main method of stunning/slaughtering chickens in the UK (DEFRA, 2022). This followed a transition away from inverting and shackling chickens by their feet to proceed towards a water bath electrical stunning system. Whilst a move away from inverting/shackling the chickens is welcome, it is far from clear if distress caused by CO2 is actually significantly less than the distress from handling, inversion, and shackling. High concentration CO2 is aversive to chickens. Their tracheal chemoreceptors facilitate perception of this as an irritant when above certain concentrations (Rucinque et al., 2024;p. 2). Further aversion can result from a sense of air hunger or breathlessness, hypoxia (low blood oxygen), hypercapnia (high blood CO2), subsequent acidosis of blood and bodily tissues, and nervous system dysfunction (AVMA, 2024, pp. 30-31;Rucinque et al., 2024). Chickens can demonstrate an awareness of CO2 at concentrations even lower than 5% (Berg et al., 2014;p. 452;Gerritzen et al., 2007, p. 181). When they cannot escape from the gas, distress behaviours commence from around 6% and can continue up to 30% concentrations (assuming incremental increases) or to the point of unconsciousnesswhichever comes first (RSPCA, 2024;Gerritzen et al., 2007, p. 182). Distress behaviours can include gasping, headshaking, high-pitched/shrieking vocalisations, jumping, and manic/strenuous escape attempts (EFSA, 2019, pp. 59, 72;Gerritzen et al., 2007, p. 182). Loss of posture then follows whereby heads are down/limp and chickens can no longer stand (Rucinque et al., 2024;p. 7). Loss of posture is often used as a proxy for loss of consciousness (EFSA, 2019;p. 27). Any of the aforementioned behaviours evident post loss of posture are not typically deemed a welfare concern as the chickens are normally considered unconscious. For chicken stunning/slaughter, the UK legal requirement is for exposure to a minimum concentration of 40% CO2. A biphasic approach to CO2 gassing refers to a first phase comprising a lower, and a second phase comprising a higher concentration of CO2. However, there is neither a requirement for a biphasic approach nor for incremental increases to the CO2 concentration (e.g., see Welfare of Animals at the Time of Killing (England) Regulations 2015, Part 5, Point 30; Protection of Animals at the Time of Killing 2009, Annex I, Ch. 1, Table 3; DEFRA, 2023). In contrast, the RSPCA Assured label requires both a biphasic approach and incremental increases to the CO2 concentration. This means there must be a first stunning phase whereby the CO2 concentration increases gradually. Additionally, in the first phase, the CO2 concentration must never exceed an average maximum of 30% (RSPCA, 2025;p. 82). The first phase brings about loss of consciousness in chickens; the second phase ensures their death in concentrations of CO2 over 40%. Whilst concentrations of CO2 40% and over are considered particularly aversive to chickens, the downside of this more incremental approach is the longer time period required for chickens to become unconscious. Evidence suggests an average of 59.2 seconds until loss of consciousness transpires in chickens placed directly into near 40% CO2 concentrations (Rucinque et al., 2024;p. 9), whilst an average of four minutes may be required in the biphasic, incremental approach adopted by the RSPCA (Gerritzen et al., 2007, p. 183). It remains unclear whether shorter bouts of more intense pain/distress are worse or better for welfare than longer less intense bouts of pain/distress (Schuck-Paim et al., 2024).According to The Independent, the RSPCA spokesperson stated that the chickens appearing in the footage were unconscious and so not experiencing any pain (Dalton & Colley, 2025). However, as outlined, CO2 stunning is not instantaneous. This is dissimilar to the stunning processes the public is somewhat used to for other farmed animal species such as cattle and sheep. Stunning of cattle and sheep typically involves the application of a captive bolt pistol to the brain. This is instantaneous if the equipment is correctly maintained and operated (Grandin, 2023). However, in the video footage in question, the process of losing consciousness has only begun, at best, for many of the chickens in the footage, as most are only in the midst of losing posture. It is most helpful to distinguish this stage by naming it as losing balance or control as Gerritzen et al. also do (2007;p. 182). Thus, awareness and the ability to feel pain/distress will remain during this process. This includes awareness of and sensitisation to the aforementioned effects of CO2 inhalation. Moreover, debate and doubt remain over whether the culmination of this process-the full loss of posture-is even a reliable indicator of loss of consciousness (e.g., Berg et al., 2014, p. 454;EFSA, 2019, p. 27). If being lenient to the RSPCA spokesperson, their statement may apply to some of the chickens who have fully lost their posture and are no longer flailing around. Crucially however, this is not true for the majority of chickens depicted in the footage. The RSPCA spokesperson's statement seems unaligned with the RSPCA's own science department's advice (RSPCA, 2024;p. 16) and appears to ignore the experience of distress more broadly, with a narrow focus on pain. There is evidence of endogenous opioid release in times of stress in mammals and avians. However, as described by Ferdousi and Finn (2018) and Baker et al. (2019), this does not necessarily result in animals (including humans) feeling no pain or distress. Furthermore, it is stressor-specific, constituting a physiological coping mechanism in some cases of stress. The authors note that 1) stress can also elicit the opposite physiological response, thereby increasing sensitivity to pain, and 2) there are multiple other sources of distress besides pain. This indicates the importance of observing a holistic set of species-specific indicators of pain and distress (Carstens & Moberg, 2000). The harmful and non-instantaneous effects of CO2 stunning are also likewise known for a number of other species, including rats (van der Schrier et al., 2022) and pigs (Mace & Knight, 2025). Indeed, there is mounting pressure to ban the gassing of pigs with CO2 (Animal Welfare Committee, 2025).The RSPCA Assured label is intended to guarantee a higher level of welfare than the legal baseline or mainstream forms of farming and slaughter (RSPCA Assured, 2025b). However, as 1) CO2 stunning/slaughter is now the predominant method in the UK (DEFRA, 2022), and 2) there is no certainty regarding shorter bouts of more intense pain/distress being worse than longer bouts of more moderate pain/distress, it is reasonable to ask: what superior welfare for chickens at slaughter is the RSPCA Assured label guaranteeing today? The RSPCA could require a meaningful and reliable higher welfare alternative for chickens slaughtered under its label. There are higher welfare alternatives available that are feasible with use of existing infrastructure. Namely, replacing CO2 with inert gases could secure higher welfare (Rucinque et al., 2024;pp. 2, 18). This is because inert gases are believed to be non-aversive to chickens (EFSA, 2019;p. 28). Specifically, the inert gas argon is recommended as it, like CO2, is heavier than air. This means it could easily be kept contained within the gas pits using current infrastructure. However, if alternatives to CO2 are being considered for new operations without infrastructure in place, then nitrogen, as an alternative inert gas, may be a preferred option as it is slightly cheaper than argon. The RSPCA's own science department encourages use of inert gases (RSPCA, 2024; p. 16), so it seems incongruent that the RSPCA Assured label does not require this.In summary, if loss of posture is accepted as an indicator of unconsciousness, the chickens in the footage in question are either beginning to lose or in the process of losing consciousness. This is because the loss of posture is not yet complete for most chickens in the footage. Hence, it cannot be asserted they have already lost consciousness and no longer feel pain, as an RSPCA spokesperson has claimed. The RSPCA could assist in securing the replacement of CO2 stunning/slaughter with argon or nitrogen stunning/slaughter to live up to their higher welfare labelling claims. Author contributions JM: Conceptualization; Writing -original draft; Writing -review & editing. AK: Funding acquisition, Writing -review & editing.The open access publication of this article was funded by the non-profit Representing Animals (UK).The authors declare that the research was conducted in the absence of any commercial relationships that could be construed as a potential conflict of interest.The author(s) declare that no generative AI was used in the creation of this manuscript. References