There is a dual accountability accepted by top executives of large multi-business firms (MBFs) to both consistently deliver performance in the short term and cultivate scalable opportunities for future growth, which requires that they be able to implement both evolutionary and revolutionary change. Researchers have defined this capability as organizational ambidexterity (OA) and deemed it a crucial requirement for every organization. Literature in the field of organizational ambidexterity (OA) describes the need to simultaneously orchestrate attention and activities along two dimensions of ambidexterity: exploitation and exploration. This dissertation peels back the layers in MBFs to generate a deeper understanding of OA in the MBF, by answering three research questions, using a mixed method approach-qualitative study followed by quantitative study: (1) What is an appropriate specification of OA in an MBF? (2) Are there differences in exploration and exploitation across the three levels within the MBF? and (3) Is there a relationship between exploration and exploitation and MBF performance at each of the three levels within the MBF? Qualitative Study 1 was conducted in December 2017, to investigate and confirm the appropriateness of the three-level operationalization of OA in MBFs. The data comprised eleven semi-structured interviews with senior executives that serve or have served as independent board directors (IBDs), chief executive officers (CEOs) or business unit general managers (BUGMs). The executives were able to articulate the existence of and differences in exploitation and exploration roles and activities at each of the three levels. Study 2 was conducted during June 2018, designed to test five sets of hypotheses. The study used a qualified panel of senior executives to generate a representative sample of respondents in each of three roles: IBDs, CEOs, and BUGMs. Data was collected through an online survey instrument, distributed to the panelists by Qualtrics and Centiment. Completed surveys resulted in a sample of 256 MBFs. Two separate one-way ANOVAs and Tukey post hoc means comparison tests were conducted; the results find significant differences in exploration at the board, corporate and business unit levels. Likewise, findings support statistically significant differences in exploitation at the board, corporate and business unit levels. For the third research question, exploration and exploitation were individually regressed against two measures of MBF performance, using IBM SPSS 25 software. At each of the three levels (board, corporate and business unit), statistically significant relationships were found between at least one of the ambidexterity dimensions and MBF performance, providing partial support of the hypotheses.