AIM: This paper reports the second phase of a study to extend the Melbourne Assessment for use with children with neurological impairment aged 2 to 4 years. The aim was to establish if (1) children's scores on the Modified Melbourne Assessment (MMA) and the Quality of Upper Extremity Skills Test (QUEST) showed a moderate to high, positive relation, (2) children had comparable behaviours for task and time demands on both tools, and (3) scores on the MMA could discriminate between children with mild, moderate, and severe levels of upper limb impairment. METHOD: An observational study of 30 children (19 males, 11 females) with neurological impairment aged 2 to 4 years. Twenty-four children had spasticity (20 with a unilateral and four with a bilateral impairment) and two children presented with athetosis, two with ataxia, and two with hypotonia. RESULTS: A high, positive relation was found between children's scores on the MMA and the QUEST (ρ=0.90; p=0.001). The clinical use of the MMA was comparable to the QUEST. MMA scores were able to discriminate between children's levels of upper limb impairment as determined by clinicians' ratings (F(2,27) =67.76, p=0.001). INTERPRETATION: These findings suggest the MMA can be clinically useful for children as young as 2.5 years and has the advantage of being valid for use with older children. Scores from the tool can also provide therapists with a quantitative means of consistently reporting level of upper limb impairment.