Parental care, a central component of reproduction in a wide range of animal species, often involves elaborate behavioural interactions between parents and their offspring. Due to the reciprocal nature of these interactions, it has been hypothesized that parental and offspring behaviours (e.g. parental food provisioning and offspring begging) are not only target but also agent of selection. These traits are therefore expected to co-evolve, ultimately leading to co-adaptation of parent and offspring behaviours within families. However, empirical data on such parent-offspring co-adaptation are limited, particularly for wild populations. Furthermore, mean levels of behaviour (as measured in previous studies) may not adequately describe the dynamic nature of the reciprocal interplay between parents and their offspring, and instead rather the behavioural reaction norms for provisioning and begging may be co-adapted. We applied a large-scale cross-fostering study over 3 consecutive breeding seasons to investigate whether provisioning behaviour of wild blue tit (Cyanistes caeruleus) parents co-varies with the begging behaviour of their genetic, cross-fostered offspring. We simultaneously analysed parent and offspring behaviours, both as static traits (mean levels) and behavioural reaction norms (offspring begging as a function of food deprivation and parental provisioning as a function of short-term experimental changes in brood size). Neither maternal nor paternal provisioning rates co-varied with the begging intensity of their genetic offspring when analysed as mean levels of behaviour. However, the slopes of the reaction norms for provisioning and begging were negatively correlated between male, but not female, parents and their genetic offspring. Thus, fathers that change their provisioning rate strongly with brood size sire offspring whose level of begging only weakly increases with hunger, and vice versa. The observed co-variation suggests the existence of sex-specific optima for parent-offspring trait combinations. Thus, our study not only highlights the importance of a behavioural reaction norm approach when investigating parent-offspring interactions, but also stresses the relevance of considering parents as separate units, at least for biparental species.