Parental investment and sexually-selected signals can be intimately related, either because the signals indicate the amount of investment that an individual is prepared to make, and hence its value as a mate (the ‘good parent process’), or because individuals are selected to vary their own investment in relation to their mate’s signals (‘differential allocation’ or ‘reproductive compensation’). Correlations between parental investment and the sexually selected signals of both an individual and its mate are therefore of central interest in sexual selection. Blue tits Cyanistes caeruleus are an ideal study species to investigate such correlations because they provide substantial amounts of biparental care and possess sexually-selected structural UV coloration that seems to signal attractiveness in both sexes. We investigated whether feeding rates of male and female blue tits were correlated with either their own or their mate’s UV coloration, and whether any such correlation was affected by the sex ratio of the brood. We also investigated whether any such correlations were reflected in offspring phenotype. Feeding rates were not correlated with either sex of parent’s own UV coloration. However, they were correlated with the mate’s UV coloration, but in opposite directions in males and females: females had higher feeding rates when mated to bright UV males, implying differential allocation, while males had lower feeding rates when mated to bright UV females, implying reproductive compensation. These relationships were unaffected by the sex ratio of the brood. In addition, fledgling tarsus length, but not mass, was related to male UV coloration, and to female UV coloration in interaction with male age. These results suggest that both male and female attractiveness influence parental investment of the mate, and that this in turn affects offspring phenotype. We found no evidence for differential sex allocation.