There is a dearth of research focusing on fathering in families of color. The present study argues that ecological factors, especially SES and neighborhood quality, exert a strong influence on racial and ethnic differences in fathering role identity, which in turn affect fathering role performance. The primary goal of the present study is thus to investigate the impact of ecological factors on what it means to be a good father among African American (n = 308), Latino American (n = 598), Asian American (n = 580), and white fathers (n = 2813) by using a nationally representative sample from the Early Childhood Longitudinal Study, Birth Cohort (ECLS-B), and to test identity theory by examining fathering identity as a primary determinant of fathering role performance. The core premise of identity theory is that society is the main source in shaping self (i.e., identity), and in turn, contributes to the way people behave (Stryker, 1968). The present study tested identity theory by examining the associations between domain-level psychological centralities and domain-specific fathering performances, and also to test whether effects of psychological centralities and contextual factors override those of race and ethnicity. Overall, the results from this study considerably buttressed identity theory. Consistent with the cultural-ecological model (Ogbu, 1981), which posits that ecological conditions shapes culture-specific socialization goals, racial and ethnic differences in the fathering psychological centrality were found because fathers in the same group historically share similar circumstances. However, the heterogeneity of the psychological centrality within each group was remarkable because their current conditions are vastly multifarious. Specifically, the lower their SES, the more likely that they believe that providing for their children is central to their identity as a father. In studying fathers of color, previous approaches often resulted in the unwitting spread of stereotypical images by contrasting minority fathers from at-risk population with middle-class white fathers, because such approaches failed to consider the effects of contextual factors on fathering and to include multiple forms of father involvement. The results from this study clearly show that racial and ethnic differences are subtle once contextual factors are taken into account. / text