This longitudinal study explored the bidirectional longitudinal relationships and sex differences in negative expressiveness among parents and internalizing/externalizing problems in their offspring. A total of 185 father-mother dyads from China participated in this study. Fathers and mothers independently completed three rounds of self-report questionnaires assessing family negative expressiveness, while mothers also reported on their offspring’s problem behaviors at the same intervals. The mean age of infants at the initial participation was 1.23 years (SD = .13). The results revealed that offspring internalizing problems at Time 1 (T1) positively predicted maternal negative expressiveness at Time 2 (T2), and maternal negative expressiveness at T1 positively predicted offspring externalizing problems at T2. Furthermore, paternal negative expressiveness at T2 mediated the relationship between early offspring externalizing problems at T1 and later internalizing problems at Time 3 (T3). The cross-lagged model also uncovered sex-specific patterns: paternal and maternal negative expressiveness predicted subsequent problem behaviors in boys, whereas problem behaviors in girls predicted paternal and maternal negative expressiveness. Notably, paternal negative expressiveness at T2 was a significant mediator for the continuity of externalizing problems in girls from T1 to T3, indicating a dynamic vicious circle with the family system. These findings highlight that the bidirectional longitudinal relationships between parents’ negative expressiveness and offspring’s problem behaviors are closely related to the sex of both parents and their offspring. Specifically, parents’ negative expressiveness exacerbated offspring’s problem behaviors (especially for boys), and offspring’s problem behaviors exacerbated parents’ negative expressiveness (especially for girls). Future interventions targeting parents’ negative expressiveness and offspring’s problem behaviors should consider the roles of sex differences among parents and infants.