INTRODUCTION In development of bioecological model of human development, Bronfenbrenner (1994:1645) and Bronfenbrenner and Ceci (1994) proposed that family learning environments can be conceived of as set of nested structures each inside other like set of Russian dolls. The proximal or immediate settings (Microsystems) in which parent-child interactions occur and linkages between those settings (mesosystems), are considered to be embedded in more remote or distal environmental contexts (macrosystems). It is proposed that to explain variations in developmental it is necessary to understand relationships among distal contexts such as family social status, proximal settings such as parent-child interactions, and measures of those outcomes. Bronfenbrenner and Ceci (1994: 572) claimed: The form, power, content, and directions of proximal processes affecting development vary systematically as joint function of characteristics of developing person, of environment - both immediate and more remote - in which processes are taking place, and nature of developmental outcome under consideration. In addition, Ceci et al. (1997: 311) stated that the efficacy of proximal process is determined to large degree by distal environmental resources.... Proximal processes are engines that actually drive outcome but only if distal resources can be imported into process to make it effective. In this study Bronfenbrenner and Ceci framework was adopted to construct and test family model to examine relationships between family environments and students' selfconcept. We chose self-concept as outcome measure as there is growing realization that positive self-concept is related to students' aspirations, completion of high school, engagement with schooling, and college attendance. The possible relationship between learning environments and development of self-concept is reflected in definition by Marsh et al. (1995: 71) that self-concept is a person's self-perceptions, formed through experience with interpretations of one's environment. The self-perceptions are especially influenced by evaluations of significant others, reinforcements, and attributions of one's behavior and accomplishments. A Family Model In constructing family model we used orientations proposed by Coleman (1990, 1997), Darling and Steinberg (1993), and by Downey et al. (1999). Coleman suggested that family influences are separable into components such as human and social capital. capital provides parents with opportunity of creating supportive proximal learning settings and it can be measured by indicators of family socioeconomic status. As Caspi et al. (1998; 427) suggested, Human capital refers to resources, qualifications, skills, and knowledge that are available to and acquired by individuals. In contrast, family social capital is defined in terms of resources that individuals may access through social ties. It is amount and quality of academicallyoriented interaction between parents and children which provides children with access to parents' human capital. Coleman (1997) suggests that if human capital possessed by parents is not complemented by strong and positive relations in families then it is irrelevant to children's educational whether parents have great deal, or small amount, of human capital. In further conceptual orientation relating to family capital, Darling and Steinberg (1993) proposed that to understand relations among distal or remote family contexts, family social capital, and students' it is desirable to examine three aspects of parenting: style, parents' expectations for their children, and practices parents use to assist children satisfy their expectations. They suggested that parenting style can best be thought of as contextual variable that moderates relationship between specific practices and specific developmental outcomes (1993: 493). …