Preschool children with delayed development of communication perform poorly on measures of speech, language and/or literacy. On standard tests of language and literacy, most are at the bottom of, or just below, the normal range, with qualitative phonological analyses showing errors typical of a younger child (Dodd, 2005a). It is not surprising, then, that children with delayed phonology are the largest group of children referred to speech language pathologists (around 60% of children with speech difficulties, Broomfield & Dodd, 2004). Recent evidence from a large community cohort study suggests that 57% of children with phonological delay, at age 4, had resolved by 7 years (Morgan et al., 2017), unlike children with disordered phonological development, whose difficulties are more likely to persist. Children with delayed communication pose a problem for paediatric speech-language pathology services. They require significant resources because of their number, but deciding whether to intervene is not straightforward. Bishop and Leonard (2001) argued that therapy for children with speech disorders should be postponed, given that so many of them spontaneously resolve. If all children with speech difficulties were left untreated, those with articulation and phonological disorders, as well as the 43% with delay who don't resolve, would accrue the disadvantages of impaired communication (McCormack et al., 2010). One compromise is to provide therapy that relies on agents of intervention to deliver cost effective group programmes in schools, given evidence (Broomfield & Dodd, 2005) that delay responds best to therapy at around 5 years of age.