This study compared different approaches for monitoring progress towards Sustainable Development Goal Target 4.2, which focuses on the proportion of children who are developmentally on track. UNICEF’s Early Childhood Development Index 2030 (ECDI2030), a parent report measure, was compared with a corresponding direct assessment measure using a sample of children aged 3 and 5 (N = 309; 154 girls) in China at two time points. In the second wave, the study also investigated the correlations and agreement between the ECDI2030 and a teacher report measure for children’s development. Although Cronbach’s alpha indicated that both adult report measures had lower reliability, McDonald’s omega showed comparable reliability among the three measures when the assumption of tau-equivalence was relaxed. Moreover, both adult report measures tended to overestimate children’s developmental levels, and were less effective in capturing the development of older children compared to the direct assessment measure. The correlations between parent report and direct assessment were significant for both girls and boys, urban children, and children from higher socioeconomic quartiles in both waves. Parents’ education levels did not substantially moderate the correlations. Moreover, parent report may not predict children’s development as effectively as direct assessment. Compared to parent report, teacher report was less effective in differentiating children’s development across socioeconomic status and urbanicity. Parent and teacher judgements were more consistent on children’s early learning competencies than on children’s motor and psychosocial skills. Implications of the findings for population-based measurement of early childhood development are discussed.