TPH2, the rate-limiting enzyme in the synthesis of serotonin, has been connected to several psychiatric outcomes. Its allelic variant, rs4570625, has been found to relate to individual differences in cognitive and emotion regulation during infancy with T-carriers of rs4570625 showing a relatively heightened attention bias for fearful faces. A significant gene-environment interaction was also reported with the T-carriers of mothers with depressive symptoms showing the highest fear bias. We investigated these associations in a sample of 8-month old infants (N = 330), whose mothers were prescreened for low/high levels of prenatal depressive and/or anxiety symptoms. Attention disengagement from emotional faces (neutral, happy, fearful, and phase-scrambled control faces) to distractors was assessed with eye tracking and an overlap paradigm. Maternal depressive symptoms were assessed at several time points during pregnancy and postpartum. The mean levels of symptoms at six months postpartum and the trajectories of symptoms from early pregnancy until six months postpartum were used in the analyses (N = 274). No main effect of the rs4570625 genotype on attention disengagement was found. The difference in fear bias between the genotypes was significant but in an opposite direction compared to a previous study. The results regarding the interaction of the genotype and maternal depression were not in accordance with the previous studies. These results show inconsistencies in the effects of the rs4570625 genotype on attention biases in separate samples of infants from the same population with only slight differences in age.