Testing of second language pragmatics is a relatively new area of language testing but has seen significant growth and development in the past 25 years. Two main streams can be distinguished. The earlier one relies on an atomistic construct of pragmatics and primarily assesses learners' ability to produce and understand speech acts, with later work also integrating other aspects, such as routine formulas and implicature. Learners' ability to use language appropriately in terms of politeness, appropriateness, and situational fit is the main assessment target. A more recent stream takes a more holistic view of pragmatics and assesses learners' interactional abilities through tasks requiring extended conversations, such as role plays and elicited conversations. The targets of assessment are learners' ability to create intersubjectivity in extended talk, perform social actions, and demonstrate sensitivity to social context. While significant research work has taken place in these two traditions, pragmatics is still not formally assessed in large‐scale language tests, probably due to the low practicality of pragmatics assessments, the challenges of creating items difficult enough for high‐proficiency test takers, and the need to attain standardization in interactionally focused tests. A significant amount of research remains to be done in these areas, with testing of pragmatics in languages other than English and the opportunities offered by intelligent agent systems being of particular importance.