The Knowledge of Infant Development Inventory (KIDI) has been culturally adapted and translated for use with populations worldwide. This study aimed to (a) test the hypothesis of the unidimensionality of the Arabic KIDI, (b) report its internal consistency, and (c) compare the differences between participants' means and variances on the latent variable, parental knowledge of child development. Two groups of participants (n = 2,195) completed the KIDI: mothers of typically developing children (Group₁ , n₁ = 1,369) and children with disabilities (Group₂, n₂ = 826). The results showed that the Arabic KIDI had a unidimensional factor structure (i.e., parental knowledge) in the two groups, with a higher single-factor model (i.e., governorates of residence). The results of the two-level factor model were associated with tenable goodness-of-fit indices. The Cronbach's alpha coefficient was adequate in each group: Group₁: 𝛼 =.79, with a 95% confidence interval of [.77,.80], and Group₂: 𝛼 = .85, with a 95% confidence interval of [.83,.87]. In addition, Group₁ had a higher mean than Group₂ on the studied underlying dimension, while Group₁ had a significantly lower variance than Group₂. The study findings implied that parental knowledge may differ due to children's disability status and place of residency. Additionally, the findings highlighted the possibility of assessing overall parental knowledge with the Arabic KIDI.