Theories on social-emotional information processing suggest that facial expression recognition (FER) is an important early step in understanding other people's feelings and intentions, enabling children to appropriately navigate social situations (Camras & Halberstadt, 2017; Crick & Dodge, 1994; Lemerise & Arsenio, 2000). Empirical studies on children’s FER indicate that by middle childhood, children typically demonstrate high accuracy in recognizing happiness while anger and sadness are more challenging, and fear is the most difficult to identify, although the order of difficulty for negative emotions may vary slightly across studies (Garcia & Tully, 2020; Lawrence et al., 2015; Staff et al., 2021). There is limited research on the development of recognition biases, which refer to the tendency to misinterpret facial expressions as certain emotions when a different emotion is actually being expressed. However, preliminary evidence suggests that in middle childhood, children often show a bias towards happiness (Garcia & Tully, 2020). To date, FER research has primarily employed a variable-oriented approach, which concentrates on changes in individual variables during development while assuming homogeneity within the population. However, studies on other aspects of emotional development and social cognition indicate, that implementing person-oriented approach might provide a more nuanced understanding of individual differences (Denham et al., 2013; e.g., Dixon-Gordon et al., 2015). Adopting a person-oriented approach in FER research would enable the identification of distinct FER profiles characterized by differences in accuracy and biases during middle childhood, as well as the investigation of stability or changes in these profiles over time. Considering the broader context of social-emotional development is essential when examining FER development. For instance, difficulties in accurately recognizing other’s emotions and biases towards specific emotions, such as anger or sadness, have shown to associate with children's social-emotional functioning, such as peer relationship problems, emotional problems, and behavioral problems (Ciucci et al., 2014; Goodfellow & Nowicki, 2009; Schepman et al., 2011; Wells et al., 2020). Furthermore, the family environment is likely to play a crucial role in FER development. For example, negative parental comments and hostile attitudes have been linked to lower accuracy in recognizing anger and sadness, while greater parental warmth is associated with higher accuracy in recognizing sadness and a reduced bias towards anger (Burley et al., 2022; Paine et al., 2021). These findings highlight the need to consider both individual and environmental factors in understanding children's FER development. While several studies have found significant relationships between children’s FER, social-emotional functioning, and family environment, these findings are not consistently observed across the literature (e.g., Dunn et al., 2018). Also, the effect sizes observed in previous research have been relatively modest (Cooper et al., 2020; Nowicki et al., 2019). One possible reason for these ambiguous findings is the predominant use of a variable-oriented approach, which assumes population homogeneity regarding these associations and may underestimate their effects (Magnusson, 2003). Therefore, in addition to the FER profiles, person-oriented approach might also provide further insight into the associations between FER development and individual and environmental factors. Taken together, the aim of the present study is to address the following research questions: (1) What are the different FER profiles in middle childhood, and how do they differ in terms of accuracy and biases? (2) How stable these FER profiles are across early school years? (3) Are these FER profiles (and potential changes in them) associated with children’s social-emotional functioning and family environment? Answering these questions would contribute to a more comprehensive understanding of children's social-emotional development.