This paper examines the challenges in allocating budgets sufficient to support rigorous evaluation of Social Innovation Fund projects. It concludes that ratios commonly used for evaluation – such as five to ten percent of program budgets – may not be adequate to meet the level of evidence and rigor required for a comprehensive evaluation. Experimental studies tend to be the most expensive; however, a variety of factors may affect these budgets, including the number of project sites and the targeted level of evidence for the evaluation. This paper includes reflections from grantees and provides a detailed snapshot of budgets for experimental, quasi-experimental, and non-experimental studies. The authors also examine budget trends for studies based on whether they seek preliminary, moderate, and strong evidence of success and the program and design factors that influence evaluation budget estimates.