Automatic coding (e.g., scoring) of short-text responses constitutes a promising method for tackling challenges associated with the constructed-response format (e.g., in adaptive testing). At the same time, it opens up plenty of new opportunities. This chapter lists such opportunities along with challenges and contextualizes the task of automatic coding. It elaborates on the functioning of ReCo, a software package for automatically coding PISA responses. An empirical demonstration shows that, once trained, the system can be reused for most items in repeated measurements without substantial effort. The chapter concludes with an outlook on how ReCo could be blended with PISA’s Machine-Supported Coding System and how it could help to harmonize human codes across different test languages.