AIM-This article explores the implementation and use of the Addiction Severity Index in addiction treatment practice, both as a clinical instrument and as a way of facilitating outcome measurement. This is regarded as incorporating "laboratory logic" into clinical practice characterised by "the logic of care". DATA-The data is based on ethnographic fieldwork in a Swedish metropolitan social service agency known for its systematic ASI work. RESULTS-The findings suggest that much effort must be dedicated to coordinate activities in the agency in line with the laboratory logic, making sure that the interviews are administered systematically. In use, the ASI and the variables in clinical practice are adjusted to each other, making it possible to follow both logics at the same time. In some cases, however, there is a conflict: the ASI becomes an extra task that does not further the clinical work. Once collected, the ASI data must be coordinated in line with other information. This has not yet been realised in the agency, which makes the value of the ASI data unknown. CONCLUSIONS-It requires hard work to handle the two logics simultaneously in addiction treatment practice: activities must be coordinated , and instruments and variables in clinical practice must be continuously "tinkered" with. Further, outcome measurement is not only about systematic use of standardised instruments, but much work must be done after the ASI data has been collected. KEY WORDS-Addiction severity index (ASI), standardised assessment instruments, outcome measurement, clinical practice, ethnography, Sweden