This paper has the aim of reinvigorating interest in the collection of detailed longitudinal data of children’s speech development. Our primary motivation is clinical. We feel that a better understanding of the trajectory of both typical and disordered development, where consonant and vowel systems are given equal consideration, would inform prognosis and more effective, efficient intervention. We begin our argument by outlining the information provided by the current approach to clinical assessment. We then examine additional benefits to be gained by using a contrastive, system wide analysis to capture the dynamic interplay between error patterns themselves and potential linguistic and environmental constraints. Note that in our description of speech examples we hold to the usual conventions whereby the orthography of the word is presented in inverted commas, the target pronunciation and speaker’s actual utterance are phonetically transcribed and presented within slanted brackets // and square brackets  respectively. We hope that the examples used mean the points of the argument are readily accessible eg ‘computer’, /k?mput?/, [put?]. We have also used examples of speech reflecting either standard southern English or standard Scottish as the target accent. We have though selected error patterns which focus on areas of commonality across any English accent rather than those on which the more distinctive features rest.