Understanding the ‘predominantly positive mental states’ required for positive animal welfare (PAW) in their complexity requires combining behavioural, cognitive and neurophysiological indicators. Regarding cognitive indicators, major gaps persist in cognitive bias research in farm animals: in an ongoing scoping review, we identified 152 studies, mainly on sheep (40), pigs (36), and chickens (33), which together account for over 70% of the identified literature. Species such as goats, donkeys, farmed fish, and other poultry species remain notably underrepresented in cognitive bias research in vertebrates. Methodologically, most research relied on judgment bias tasks, which provide insight into animal optimism or pessimism but capture only part of the cognitive landscape. Attention biases (focused on emotionally salient stimuli) are gaining ground but remain relatively underexplored. In addition, memory bias tasks, which assess how past emotional experiences shape information processing and recall, have never been studied in farm animals. To address these limitations, we also consider human literature to identify standardised cognitive bias testing methods that could be adapted for animal research. This widened approach holds promise for diversifying methods, expanding species focus and furthering our understanding of PAW across cognitive and emotional dimensions. While cognitive indicators reflect how an individual’s brain can be biased due to their current affective constitution, specific mental states are manifested through brain changes measurable at different levels of the body. On the one hand, peripheral measures relevant for PAW include, among others, heart rate variability, plasma neurotransmitters (e.g. serotonin, dopamine) and neurohormones (e.g. oxytocin, vasopressin). These measures enable assessing affective reactions in awake and free-moving animals in emotionally relevant contexts. This is critical to develop methods that can eventually be applied on-farm. On the other hand, the rise of studies directly measuring brain activity (central measures) in farmed animals has the potential to more directly assess positive affect. Such measures include electromagnetic (e.g. EEG) and brain imaging (e.g. (f)MRI) techniques. Finally, post-mortem measures can provide evidence for short-term (e.g. immediate early gene expression) and long-term (e.g. hippocampal plasticity) positive affect. An advantage of post-mortem measures is the use of tissues collected at slaughter without needing additional invasive procedures. Using a holistic approach that combines both peripheral and central measures is a timely solution for advancing PAW and eventually applying relevant indicators in real (onfarm) conditions