The definition of animal welfare, and how to improve it, has evolved over the years. Historically, animal welfare was seen as guaranteeing basic animal needs and absence of suffering: freedom from hunger, thirst, discomfort, poor health, fear and distress, and an environment that provides animals opportunities to express their natural behaviour. Since then, animal welfare is growingly focussing towards providing opportunities for animals to have positive experiences and emotions. Although satisfying farm animals’ physiological needs can be quite straightforward, improving animals’ emotions is a tremendous scientific and practical challenge. Behavioural genetics is crucial to help improve abilities to have positive experiences at the individual and at the group level in poultry. Cage-free and free-range systems are more respectful of animals’ needs, enabling them to express natural behaviours such as exploring, perching, scratching, dust-bathing, interacting, escaping… However, alternative farming systems challenge animals who have been selected for their performance rather than their robustness or resilience. As birds are given more opportunities to express their natural behaviours, the genetic contribution in expressing these behaviours and their trade-off with production traits need addressing. In cage-free systems, a top priority is for laying hens to lay their eggs in the provided nests. We will review recent advances in understanding hens’ egg-laying behaviour in nests and show that there are some signs that nest laying could be genetically improved. Cage-free systems are also a great challenge for breeders to select their population when they freely mate on the ground. When not controlling for reproductive pairings, cage-free reproduction paves the way for uncontrolled natural selection to interact with breeders’ artificial selection. Beyond very costly genetic parental assignment of all hatched chicks, we will see how understanding roosters mating behaviour could provide keys to guarantee genetic diversity in selection lines. Finally, free-range poultry is growing in popularity but individual range use behaviour and their genetic influence nevertheless needs new tools to be deciphered. Once animals are adapted to their environment, they would still vary in their individual emotional state. By understanding animals’ perceptions, emotions, and experiences, behavioural scientists are able to investigate new traits and characterise their genetic variance, but genetic contributions to emotions are very little explored yet. We will review the genetic characterisation of three traits that could improve birds’ experience. First, optimism, a cognitive judgement bias that reveals positive expectations, showed some molecular basis suggesting genetic variance. Second, farm animals will always have to endure some stressful events. Emotional reactivity can be selected in quails, and may be part of a usable toolkit to improve poultry’s experience. Finally, feed restriction in broiler breeders is a source of negative experience, but feeding them ad libitum causes an even more deteriorated welfare. Increasing satiation in birds, even when they are only lightly fed, may help tackle this welfare issue. Although selecting on positive experiences and emotions may improve poultry’s welfare, selection on those traits can never replace providing ethical conditions to farm animals. Positive emotion traits should only be used as a supplementary tool to improve animal experiences in their lives, within an adapted environment. Finally, all poultry species have strong social structures from their wild ancestors. However, modern conditions and density vastly differ from ancestral conditions and animals develop behaviours that may harm the welfare of the herd such as feather pecking or cannibalism. The genetic background of these harmful behaviours has been notoriously difficult to demonstrate. However, the emergence of digital tools (e.g. radio-frequency tagging, image analysis, etc.) enables individual identification and phenotyping of behavioural traits, opening up new breeding prospects for these traits. Beyond trying to suppress harmful social behaviours, scientists are now interested in understanding positive emotions coming from social experiences. For example, preferential clustering has been observed in several production systems, and its effect on animal welfare requires exploring. More generally, data on behaviours that enhance animal well-being, experiences and emotions are very scarce, as behaviours generating injuries, mortality, and lower performances have been prioritised. However, if we want to move towards improved poultry welfare, we also need to gather efforts towards phenotyping and genetically characterising behaviours that improve positive experiences in the lives of our animals.