Spelling analyses can be used to investigate sources of linguistic knowledge underlying children’s literacy development and may be useful in predicting later achievement. This study explored the utility of six analysis metrics in predicting the spelling achievement of school-aged children with literacy learning difficulties via post-hoc analyses of data collected in a clinic. Participants were 48 children aged 7 to 12 years. Spelling accuracy was assessed using the Dalwood Spelling Test (Dalwood Assessment Centre, 2008) at baseline and 37–70 weeks later. Spelling attempts at baseline were analysed using metrics designed to quantify evidence of phonological, orthographic, and/or morphological awareness. Scores from each metric were associated with baseline and later conventional spelling accuracy. A metric which credits evidence of phonological, orthographic and morphological awareness shared a significantly stronger association with baseline conventional spelling accuracy as compared to the remaining metrics. There were no significant differences in the strength of associations among the baseline metrics and later spelling achievement. Supplementary analyses focused exclusively on children’s spelling errors returned a similar pattern of results with a few notable exceptions. The utility of spelling analyses is discussed.