Carbon dioxide (CO2) is commonly used for stunning and euthanasia of pigs, but the practice is restricted in some countries due to concerns about pain and distress that appear to occur prior to loss of consciousness when pigs are immersed in CO2. Argon-induced hypoxia is suggested as an alternative, but is associated with prolonged induction of unconsciousness during which pigs display behaviors suggestive of distress. The aim of this thesis was to assess pigs’ aversion to inhalant euthanasia gases using approach-avoidance (AA) and conditioned place avoidance (CPA) paradigms. A preference-testing device was designed with two identical chambers: a control chamber (CC) with ambient air and a treatment chamber (TC) with different concentrations of CO2 and O2 gases, separated by a sliding door and an exhaust sink. Weaned crossbred commercial pigs were individually trained once daily for five days to enter the TC through the sliding door to obtain food rewards, with chambers maintained at ambient air conditions. Following entry to TC, the pig could move freely between chambers. Testing involved the same methods, with TC maintained at 10%, 20%, and 30% CO2 (Chapter 2) or 6% and 2% O2 (Chapter 3). Tests concluded when loss of posture (LOP) occurred, or after the test concluded. Pigs were systematically assigned to each gas treatment over three rounds (Chapter 2) or two rounds (Chapter 3). Each test round consisted of ambient air on baseline (B) day, assigned gas treatment on gas (G) day, followed by ambient air on washout (W) day. It was hypothesized pigs would avoid gas concentrations that were aversive on G and CPA would be observed on W. Behavioral outcomes were collected using live and video recordings. In Chapter 2, 10 of 12 pigs entered the TC on all B, G and W days, followed by six minutes during which pigs could move freely between chambers. Pigs displayed longer latency to enter TC, shorter latency to leave TC and longer latency to re-enter TC on G than B days. Five pigs at 20% and 4 pigs at 30% CO2 remained in the TC until LOP. CPA was not observed on any of W. Hence, all CO2 levels tested induced mild aversion compared to ambient conditions, but this aversion did not provoke marked avoidance or CPA. In Chapter 3, all 12 pigs entered TC on all B., G. and W days, followed by 10 minutes during which pigs could move freely between chambers. Pigs displayed shorter latency to leave TC and longer latency to re-enter the TC on G than B days. At 2% O2, seven pigs remained in TC until LOP. CPA was not observed on any of W. Hence, hypoxia appears to be mildly aversive to pigs, however the degree of aversion was not pronounced to provoke marked avoidance or CPA. In conclusion, 30% CO2 and 2% O2 were not sufficiently aversive to provoke avoidance or CPP in weaned pigs, and are suitable for gradual-fill or two-step euthanasia protocols for on-farm euthanasia of pigs.