Introduction: Previous research reported transactional relations between child functioning and parenting stress. There is limited evidence whether a transactional developmental model also fits children below the age of 12 months, especially in psychosocially burdened families. This study aims to test the fit of a transactional model during the first 3 years of life and examines whether the model differs between families with low and high psychosocial burden. Methods: A total of 302 psychosocially burdened families were observed over 3 years at age 4, 12, 24, and 36 months. Child behavioral problems and parenting stress were assessed via self-report while psychosocial burden was assessed via external rating at baseline. Cross-lagged panel analysis was used to investigate the fit of a transactional model. Results: A transactional model fitted the data significantly better (Δχ<sup>2</sup> = 81.87, p < 0.001) than an autoregressive model reaching acceptable to good fit indices (CFI = 0.96, RMSEA = 0.09). The model indicated moderate stability within and reciprocal effects between child behavioral problems and parenting stress from age 12 to 36 months. From age 4 to 12 months, parenting stress predicted child behavioral problems but not vice-versa. Model fit indices and transactional relations did not substantially differ between families with low and high psychosocial burden, except for child effects on parenting stress during the first year of life, which were only evident in higher burdened families. Conclusion: Transactional relations among child and parent variables are evident in the first 3 years of life. Child effects in the first year of life may be restricted to highly psychosocially burdened families. Future research may focus on potential mediating variables such as parental sensitivity or contextual variables like significant life events. Targeted prevention strategies should be adapted to the level of psychosocial burden to account for the differing transactional relations.