Abstract
The original proposal of Goal Attainment Scaling (GAS) by Kiresuk and Sherman (1968) offered a method of evaluating services which had to set individualised goals for clients because of the diversity of their needs. The method expressed clients’ levels of achievement of their goals on a common, normative scale, thus creating a basis for quantitative evaluation. This paper recognises the attraction of GAS, but suggests that there are concerns about the calculation of its standard scores that should be added to concerns about procedures which other writers have identified. Examples are given of a different approach to the displaying of GAS data that makes fewer assumptions than does the original technique. These examples show how GAS may have a useful role in service development whether or not numerical values are attached to clients’ levels of achievement.