I want to present some ideas about how a valid and reliable process for assessing deep thinking is not a function of the assessment regime (such as external or internal, standardised or teacher-devised), but is actually a product of the successful application of certain design criteria and the interplay of three essential elements. The argument I present rests on one simple belief that I hold: the capacity to design good assessment tasks is a vital part of an extensive professional repertoire and, as such, demands space and time, ritual and respect. (Teacher–assessors should not let anybody tell them that designing assessment tasks for deep thinking is a job for a rainy Sunday afternoon.)