The TOEFL iBT test was introduced in 2005 to better reflect the language demands of real-life academic tasks than did previous versions of the test. The task-based design of the test was intended to support the interpretation of its scores as a trustworthy measure of international students' ability to use English in an academic environment. Until recently, the TOEFL iBT test was exclusively administered online at secure test centers around the world. In response to the disruption caused by the pandemic in early 2020, the TOEFL iBT Home Edition was launched, which is identical to the version administered in test centers, but taken at home through the examinee's computer, in a proctored environment. We present key features of content delivery and the security of the test. We then discuss preliminary findings related to the comparability of scores across modes of delivery, and important implications for the design and score interpretation of at-home language proficiency tests.