Introduction: Stigma refers to the possession of certain characteristics or the belief that a person possesses characteristics associated with a less valuable social identity in a society (Crocker et al., 1998). There are four interrelated types of stigma (Pryor & Reeder, 2011; as cited in Bos et al., 2013): (1) public stigma, refers to society's attitudes and reactions toward stigmatized groups, (2) structural stigma, refers to the legitimization of a stigmatized status, (3) selfstigma, refers to the internalization of public stigma, and (4) stigma by association, refers to the transmission of stigma to people close to stigmatized individuals. Stigma by association can also be internalized as affiliate stigma (Mak & Cheung, 2008). According to the sociocognitive model of self-stigma (Watson et al., 2007; Corrigan & Rao, 2012), self-stigma develops in four sequential stages. First, a person perceives the stigma in public (awareness phase), then begins to accept negative stereotypes from the public (acceptance phase), and when he or she begins to apply these stereotypes and prejudices to him or herself, internalization of the stigma occurs (application phase). Finally, internalization of the stigma leads to impairment of the person's psychological functioning (harmful consequences phase). The model has not been fully confirmed empirically (see e.g., Watson et al., 2007; Corrigan et al., 2011; Göpfert et al., 2019). Two models of self-stigma of parents of children with disabilities have been tested in previous studies: the model proposed by Eaton et al. (2020) and the attribution model (Čolić et al., 2021; Mak & Kwok, 2010). The present study also tested the attribution model of parent self-stigma, which proposes that attributions of the cause of child developmental difficulties mediate the relationship between perceived and internalized stigma. Previous research has shown that attributions of the cause of a negative event are related to one's adaptation. According to Weiner's (1985) theory, all causes can be classified on three dimensions, locus (internal vs. external causes), stability (stable vs. unstable causes), and controllability (controllable vs. uncontrollable causes). Depending on how one perceives the cause in terms of locus, stability, and controllability, different cognitive, emotional, and behavioral responses occur. The attribution approach has proven to be a useful theoretical framework for explaining stigma, although it has been explored primarily in the area of public stigma. Results consistently show that public stigma is greater when people attribute the stigmatizing condition to controllable causes (e.g., Weiner et al., 1988; Schwarzer & Weiner, 1991; Corrigan et al., 2003). The role of attributions in the process of self-stigma has been less researched, and previous findings have been inconsistent (e.g., Mak & Wu, 2006; Boyle, 2016; Mak & Kwok, 2010). Studies have primarily focused on the controllability dimension, although some research on the relationship between attributions and adjustment to negative life events suggests that other dimensions should also be considered (e.g., Boyle, 2016; Roesch & Weiner, 2001). The focus of research has only recently shifted to the self-stigmatization of parents of children with developmental disabilities. Previous studies (e.g., Mak & Cheung, 2008; Mak & Kwok, 2010; Eaton et al., 2020) have found low to moderate levels of stigma in parents, but internalized stigma has consistently been associated with negative psychological outcomes (for a review, see Papadopoulos et al., 2018). The focus of the present study was on selfstigma in parents of children diagnosed with the developmental language disorder, a population that has not been previously studied. The developmental language disorder is a condition in which children have difficulty using and/or understanding language without any other known deficits that lead to language difficulties (Leonard, 2014). It is a heterogeneous disorder with various clinical manifestations that persists into adulthood. Although the external symptoms are usually less visible, the disorder is associated with various negative consequences, both for the child (e.g., Botting & Conti-Ramsden, 2000; Conti-Ramsden & Botting, 2004; Young et al., 2002) and for the parents (e.g., Macharey & von Suchodoletz, 2008; Marshall et al., 2007). Therefore, the aim of this study was to test the attribution model of self-stigma in parents of children diagnosed with the developmental language disorder. We examined (1) whether attributions of the cause of the child's disorder mediate the relationship between perceived and internalized stigma, and (2) the effects of stigma and attributions on parents' psychological functioning. Methodology: We conducted a three-wave longitudinal study with a three-month time interval: the first wave in December 2020, the second wave in March 2021, and the third wave in June 2021. Data were collected from parents whose children had been diagnosed with the developmental language disorder. The children were between 4 and 15 years old and the majority (93 %) was receiving speech and language therapy. Parents whose child had a comorbidity with a more severe disorder or was physically disabled were excluded from the sample. Participants were recruited through speech-language pathologists and, to a lesser extent, online advertisements. The final sample consisted of 380 parents from different regions of the country: 171 (45 %) participated in all three waves, 98 in two waves (25.8 %; N12 = 67, N23 = 17, N13 = 14), and 111 in only one wave (29.2 %; N1 = 74, N2 = 17, N3 = 20). The mean age of participants was M = 39.63, SD = 5.15, and 94 % were mothers. The sample was heterogeneous in terms of time elapsed since the child's diagnosis (age range: 0-12 years, M = 3.44 years, SD = 2.39). Data were collected online via SurveyMonkey. Participants were asked to complete revised versions of the Devaluation of Consumer Families Scale (Struening et al., 2001) and the Affiliate Stigma Scale (Mak & Cheung, 2008), the Revised Causal Dimension Scale, CDS-II (McAuley et al., 1992), the Depression Anxiety Stress Scale, DASS-21 (S. H. Lovibond & P. F. Lovibond, 1995), and a sociodemographic questionnaire designed for the purposes of this study. Participants received the invitation to complete the questionnaire via their email addresses collected before the start of the study. In each wave, the invitation was sent to all addresses, regardless of whether the participant had taken part in the previous wave. At the end of the study, all participants received a short report with the main conclusions of the study and information brochures on parental stigma and developmental language disorder as a thank you for their participation. Data were analyzed using linear structural equation modeling. Facet-representative item parceling was used to create indicators for latent variables. The parcels formed as means of the subscales were used as indicators of perceived stigma, internalized stigma, and negative emotional states, and the items of the subscales of CDS-II scale were used as indicators of the locus, stability, and controllability of the cause of the child's developmental language disorder. Several nested cross-lagged models were tested using maximum likelihood with robust standard errors (MLR). Missing data were handled using full information maximum likelihood (FIML). Results: Consistent with the socio-cognitive model of self-stigma, the results showed a temporally stable effect of perceived stigma on internalized stigma over time. Higher perceived stigma was associated with an increase in internalized stigma three months later. However, contrary to our hypothesis, perceived stigma had a direct longitudinal effect on internalized stigma. The hypothesized mediating role of attributions of the cause of the child's language impairment in the relationship between perceived and internalized stigma was not confirmed. Perceived stigma had a significant longitudinal effect only on the locus of the cause of the child's disorder. Parents who perceived more blaming and devaluing in their social environment were more likely to attribute their child's disorder to internal causes (causes related to the parents) three months later. However, the effect was found only in one time interval (from T2 to T3). One possibility is that attribution effects were not obtained because participants were heterogeneous in terms of time elapsed since the child's diagnosis, i.e., parents were not at the same stage of adjustment to the child's diagnosis. The relationship between attributions and adjustment to negative events could differ over time. Therefore, it is important to consider the time since the event when interpreting attributions and their relationship to adjustment No longitudinal effect of internalized stigma on parents' negative emotional states was found either. Internalized stigma did not mediate the relationship between attributions of the cause of the child's disorder and negative emotional states, but perception of the stability of the cause of the child's disorder measured at T1 had a direct positive effect on negative emotional states measured at T2. In addition, negative emotional states at T2 had a significant positive effect on internalized stigma at T3. In other words, the results suggest that a greater perception of the stability of the cause of the disorder leads to higher levels of negative emotional states in parents, and impaired psychological functioning leads to greater internalization of stigma. However, the indirect effect of the stability of the cause of the child's disorder on internalized stigma via parents' negative emotional states was not significant, most likely due to insufficient statistical power. Although some hypothesized longitudinal relationships were not found, this does not mean that they do not exist. It is possible that the time interval was not optimal to achieve certain effects. To our knowledge, this is the first study of self-stigma in parents of children diagnosed with the developmental language disorder. The results suggest that the conclusions about the relationship between perceived stigma and internalized stigma over time may be generalizable to a population in which the disorder is less visible. Conclusion: The present study confirmed the longitudinal effect of perceived stigma on stigma internalization. The stigma was internalized to a greater extent by individuals who were more aware of the existence of the public stigma. Not all parents who perceived the stigma internalized the stigma. The mechanisms underlying the internalization of stigma are still poorly understood, and the present study did not confirm the mediating role of attributions. Although it is widely believed that internalization of stigma leads to impaired psychological functioning, our findings, as well as some previous research, suggest that this relationship may be bidirectional and more complex. We found that higher perceived stability of the cause of the child's disorder was associated with an increase in negative emotional states in parents and that parents with impaired psychological functioning were more vulnerable to internalizing stigma over time. Findings suggest that interventions to reduce stigma should aim to reduce public stigma against parents of children with developmental disabilities and educate parents about mental health care strategies. Professional psychological support should be offered to those who have difficulty coping with stress and negative feelings.