This paper starts with an analysis of the twin purposes of an examination reform programme originally launched in Kenya during the 1970s: first, to broaden the spectrum of cognitive skills being tested; and second, to set up a feedback system based on the performance profiles, providing schools and teachers with guidance as to how pedagogy and learning might be strengthened. Then, in a change of time perspective, the paper looks back over 30 years of implementation experience, examining the reasons why some components of the original programme have proved more sustainable than others. A final section discusses three key issues which have emerged in recent years, and considers their implications for policy.