This study adds clarity to inconclusive results in previous literature about the link between board characteristics and the level of social responsibility performance by using fuzzy set qualitative comparative analysis. Specifically, we propose a new holistic framework based on the complexity theory. From a sample of 471 non-financial companies from the USA for the period 2008-2010, our findings support the four tenets of complexity theory: equifinality, complexity, asymmetry, and causal asymmetry. More concretely, they suggest that CSR performance depends on a complex configuration of some board characteristics, such as size, independency, diversity and activity, and other corporate attributes (i.e., company size, leverage, and growth opportunities). These factors play a key role as the ingredients of the recipe and, in a proper combination, contribute to obtaining high levels of social responsibility performance.relationships (Wu et al., 2014). We feel the need of multidimensional complex analyses; focusing on single dominant hypotheses, such as previous researches, lacks objectivity (Armstrong et al., 2001), and ignores the complexities inherent in realities (Woodside, 2014).Accordingly, our aim is to explore combinations of board structure and other corporate characteristics that lead firms to achieve high levels of CSR performance, based on complexity theory. The core premise of this theoretical framework is that one specific outcome can result from different combinations of causal factors (Ragin, 2000). Following this idea, several combinations of board characteristics (e.g. size, independence, diversity, and activity), as well as corporate characteristics (e.g. size, growth, and leverage) could cause a high level of CSR performance.This goal is addressed for a sample of 471 US non-financial firms for the period 2008-2010. By using a qualitative comparative analysis (QCA), our results suggest complex paths, since CSR performance is explained by more than three-way interactions of board and corporate characteristics. In addition, our findings suggest that single configuration of such characteristics may be sufficient but not necessary for predicting the level of CSR performance (equifinality). We also found asymmetry situations, since a specific simple condition may contribute both high and low levels of CSR performance, as well as causal asymmetry, i.e., configurations associated with high outcomes are not the mirror opposite of combinations associated with low levels of CSR.This paper contributes to previous literature focused on CSR and board of directors by proposing a holistic approach. We note that CSR performance does not depend on individual board characteristics (e.g. size, independency, diversity, and activity) but on specific configurations of such features. Furthermore, corporate characteristics (e.g. size, growth, and leverage) and activity sector are also involved in these configurations. As far as we know, this is the first attempt to use a configural analysis framework by using fuzzy-set qual...