Background: Over the past few decades early self-regulation has been identified as foundational to positive learning and wellbeing trajectories. As a consequence, a wide range of approaches have been developed to capture children's developmental progress in self-regulation. Little is known, however, about whether and which of these are reliable indicators of future ability and risk for young children. Objective: This study examined measures from prominent approaches to self-regulation assessment (i.e., task-based, observation, adult-report) to determine: their structure; how these predict future academic school readiness in 3-5-year-old children, individually and if combined; and whether thresholds could be ascertained to reliably discriminate those children at risk of poor academic outcomes. Methods: Longitudinal analyses were conducted on start-of-year self-regulation data from 217 children in the final year of pre-school, using three prominent approaches to self-regulation assessment, and their end-of-year school readiness data. Data were subjected to path analysis, structural equation modelling and receiver operating characteristic curve analyses. Results: Start-of-year cognitive self-regulation indices--but not behavioral or emotional self-regulation indices--from each approach reliably predicted school readiness 7 months later, just prior to commencing school. Only when combined into a composite score was a threshold with sufficient sensitivity and specificity for predicting school readiness risk established; yet this provided better prediction of true-negative than true-positive cases. Conclusions: Taken together, these results suggest the importance of cognitive self-regulation in particular for school readiness, as measured here, although self-regulation is just one of the contributing factors to school readiness risk.