The new options that technology offers for language test tasks present teachers, test developers, and researchers with an expanded set of possibilities for assessing language ability. At the same time, however, these options raise questions about what the new tests actually measure and what test scores from CALT can be used for. In our experience, the mention of technology for language assessment prompts skeptical comments from teachers, test developers, and researchers. One language-testing researcher characterized the skepticism by writing about the potential threats of computer-assisted language testing in addition to the promises. Expanding on the promises, Chapter 2 discussed the new options afforded by computer technology within the framework of the test method characteristics. This chapter examines the threats by making explicit connections between concerns about technology in language testing and validity. Some concerns about CALT that we mentioned in Chapter 2 pertain to (1) the inferences that can be drawn from test scores on computer-based tests and (2) the uses of the scores for purposes such as certification, admissions decisions, assigning grades, and advising learners. These two validity issues form the basis of at least six concerns that are often expressed as potential threats to CALT. In this chapter, we identify these potential threats and explain the ways in which research might address them. The fact that this chapter raises more questions than answers reflects the status of knowledge about inferences and uses of CALT, but identification of these issues is a first step to progress in this area.