Diaries, as a tool for data collection, have been around for some time. Lessons shared to date come from disparate settings and there remains a degree of ambiguity regarding the value of diaries, particularly in resource-poor settings where populations are often illiterate and highly mobile. We recently designed a pictorial diary for the collection of data on household consumption and expenditure in Tanzania and The Gambia. A random sample of 361 diary keepers in The Gambia and 308 in Tanzania maintained diaries for a period of 12 months. The aim of this paper is to share some of the lessons learnt in developing and applying this instrument. It is structured around a series of questions about diaries that we found relatively few answers to when we first embarked on this study. These questions include: how should a diary be designed? How long should a diary be maintained? When should entries be recorded? Who should keep the diary? The motive behind this paper is simple: to provide future researchers who are contemplating using diaries in resource-poor settings with some practical information that may guide them through this process.