Background: Many children referred to the child welfare services are not screened in for further investigation. Factors related to intake decisions have been found on several levels. Many studies, however, lack information about the children's own experiences of child maltreatment and/or behavioral problems. Objective: To examine case factors relating to decision to investigate referrals to the child welfare services in Sweden. Participants and setting: Data are used from a prospective longitudinal multisource program (LoRDYA) in Sweden studying two cohorts of adolescent children in four municipalities (n = 1884). Methods: By linking annual self-rating data with registry data from the child welfare services, data are analyzed through latent-class analysis and Poisson regression. Results: Most children who are self-rated severely exposed to maltreatment and/or behavioral problems are never investigated by the child welfare services (74.2 %). Referrals concerning girls are more likely to be investigated than referrals concerning boys (crudeAME = 0.09 p = ***). For girls, prior referrals (adjAME = 0.16, p = ***), household poverty (adjAME = 0.09, p = *) and any form of self-rated severe exposure to maltreatment and/or behavioral problems (adjAME = 0.14, p = ***) increased the chance of a decision to investigate. For boys, prior referrals (adjAME = 0.24, p = ***) increased the chance of a decision to investigate, while achieved age (adjAME = −0.03, p = *), and being referred on suspicions of neglect and behavioral problems (adjAME = −0.15, p = **) decreased the chance of a decision to investigate. Conclusions: Child welfare services assess referrals on boys and girls differently, which may explain why boys have in crude numbers a decreased chance of decisions to investigate.