In this talk, I will present recent work examining deep subcortical circuits and their interactions with prefrontal cortex during affective decision-making. I will focus on three complementary strands of research that together inform our understanding of these networks.First, I will describe progress in resolving small subcortical structures that are critical for emotional and motivational processes—such as the amygdala and hypothalamus—at the level of individual subnuclei. Using high-resolution neuroimaging, we show nucleus-specific patterns of brain connectivity that explain variance in mental well-being, including individual differences in negative affect and stress.Second, I will discuss advances in causal approaches that allow us to move beyond correlational descriptions of function and directly manipulate activity non-invasively in deep brain circuits. Using transcranial ultrasound stimulation, we have characterised the causal contribution of the basolateral amygdala to affective approach–avoidance decisions, and ongoing work extends this work to examine causal roles of prefrontal and striatal regions in affective behaviour.Third, across multiple domains, we are beginning to move beyond the timescale of individual trials to study intermediate, more naturalistic timescales. This work aims to characterise how background contextual features shape motivation, social behaviour, reward learning, and emotion processing. Taken together, our work in human cognitive neuroscience has moved from correlational studies with coarser anatomical and functional resolution toward causal investigations of deep subcortical-cortical brain circuits at the functional scale of individual nuclei and using timescales of increasing relevance for flexible human behaviour. We believe this is an important step towards understanding the neural mechanisms underlying affective decision-making across health and disease.