Firms recognise agility as a crucial capability to survive and thrive in uncertain and turbulent markets. One of the reported enablers of agility, from an IT perspective, is the Service-Oriented Architecture (SOA). SOA with reliance on its characteristics has provided faster time to market and reduced system complexity. Review of SOA empirical results, however, suggested an increased process and system complexities which conversely inhibit achieving the goal of agility. The contradictory effect of SOA on complexity and lack of insight on IT sensing capability motivated this research to study the effects of SOA on the two-underlying sensing and responding capabilities of the IT agility. Due to the current limited theoretical perspectives in the extant SOA studies, the current study undertakes a theory-building research. With reliance on the real options theory, complemented by the dynamic capabilities, the current study develops an initial theoretical framework and argues that SOA characteristics, when embedded in a system at the design time, will provide future knowledge and process options. The embedded future options, when executed, will facilitate the IT sensing and responding capabilities. By collecting and analysing data from twenty-two in-depth interviews as well as project documents across multiple cases in a Bank, an Airline company, and an Airport, this study extends its initial conceptual framework to a mid-range theory that explains the interaction between SOA and IT agility. Results of the study contribute to the SOA literature by conceptualising the SOA in three sets of characteristics including: ‘information-centric’, flexibility-centric’ and ‘structure-centric’, as well as theorising the effects of each set on the effectiveness of process options and ‘change detection’ and ‘shared insight’ knowledge options. Furthermore, the study took steps to extend the options theory by developing a new concept called options depreciators, which represents the negative effect of options when there are multiple competing options. Finally, this dissertation contributes to the SOA governance literature by providing a novel view on the required governance structure for achieving agility. Under this perspective, this study presents an adaptive governance structure that is consistent with the SOA hierarchical layering and also appropriate to foster business autonomy and to create sustainable assets for future use.