This study seeks to better understand the factors contributing to the adoption of Software as a Service (SaaS) as on-demand sourcing option. Grounded in transaction cost theory (TCT), we developed a research model for assessing SaaS-sourcing at the application level. Four hypotheses with three TCTbased constructs (application specificity, environmental uncertainty, and usage frequency) and one contingency factor (firm size) affecting the adoption of SaaS-based sourcing were formulated. Survey data of 154 firms in Europe with 284 response items across different industries were collected to test the theoretical model. PLS-based structural equation modeling demonstrated that uncertainty emerges as the strongest factor for SaaS-adoption, while application specificity also contributes significantly. Differentiating between small and medium-sized businesses (SMBs) and enterprises, uncertainty represents the strongest driver for SaaS-adoption in SMBs, whereas application specificity is more important in enterprises. Finally, firm size is significantly and negatively associated with SaaS-based sourcing within, but not across the subgroups of SMBs and enterprises.