Natural selection acts on underlying mechanisms to fine-tune expression of behaviour across scales – within individuals, among individuals, between sexes and across species. The inherently environmentally responsive nature of behaviour, or behavioural plasticity, may bias how behaviour evolves. However, studies of plasticity rarely integrate patterns across levels of biological organization and timescales, limiting our understanding of how mechanisms of individual behavioural variation translate to patterns of evolutionary diversification. Here, we advocate that the contributions of plasticity to evolution between populations and species cannot be fully understood without consideration of plasticity within and between individuals; particularly in view of the often assumed but rarely tested idea that mechanisms mediating plasticity are the same as those targeted by selection. Using parental care as a touchstone, we explore how mechanisms involved in behavioural plasticity in individuals may (or may not) be co-opted to generate behavioural variation among individuals, between sexes and across species. We draw on parental care diversity in poison frogs to explore how different patterns across levels of biological organization inform our predictions of evolutionary outcomes and advocate for more empirical studies of proximate mechanisms underlying individual variation. Alongside a renewed appetite for empirical studies of phenotypic plasticity , recent years have seen major technological advances that have made mechanistic studies increasingly possible, even in nonmodel systems. Parental care is only one example, and we emphasize that these concepts apply to the general conversation about the evolution of behaviour, behavioural plasticity and phenotypic plasticity more generally.