Motor difficulties associated with Developmental Coordination Disorder (DCD) are frequently apparent before the accepted diagnostic age of 5. Tools to support identification of DCD markers would allow provision of early intervention to reduce negative sequelae. Establish psychometric properties and define preliminary cut-off scores for the Brazilian Little Developmental Coordination Disorder Questionnaire - Brazil (LDCDQ-BR). Parents of 3- and 4-year-old children (n = 312; 154 girls) from Belo Horizonte/MG, Brazil, completed the LDCDQ-BR, the Brazil Economic Classification Criterion and a demographic questionnaire. One sub-set of children (n = 119) was assessed with the Movement Assessment Battery for Children–2nd Edition; another sub-set (n = 77) completed the LDCDQ-BR a second time. Rasch analysis indicated good item functioning with only one erratic item, suggesting unidimensionality. Item calibration reliability was excellent (0.97), children’s measures reliability was low (0.72), but implying separation of 2.46 motor ability levels. Significant, low correlations were found between the LDCDQ-BR and MABC-2 (r = 0.30, p < 0.01). Test-retest reliability was 0.77 (total score) and 0.44–0.78 (individual items). ROC curve analysis revealed sensitivity of 68% at a cut-off score of 64. The LDCDQ-BR shows promising psychometric properties to support early identification of DCD. •Rasch analysis was used to examine the measurement properties of the LDCDQ-BR.•Items were easy, but fit the Rasch model, suggesting unidimensionality.•Total score was valid, but subcategory scores were not.•The items discriminate two levels of motor ability.•Total score ≤ 64 suggests possible DCD in 3 and 4 years old Brazilian children.