Psychometric network analysis (PNA) is a developing approach to explore relationships among variables. Unlike theory-driven techniques (e.g., path analysis), which assume directional relationships on theoretical models, PNA is a data-driven approach that can help to identify emerging relationships among variables without assuming a specific structure. Moreover, within PNA it is also possible to compute measures that indicate the variables with the greatest influence, and to assess stability and replicability of estimated networks and measures. Based on these premises, in this study, we reanalysed data of a previously published research aimed at identifying the role of cognitive predictors of drawing skills, analysed with path analysis, using PNA to compare the two approaches on real data. The considered sample included 142 children (72 boys, 70 girls), aged 7 to 11 years (M = 8.8, SD = 1.1), who underwent a neuropsychological assessment for: visual attention, visual perception, complex spatial abilities, visual-motor integration, working memory (verbal and visuospatial), verbal abilities and drawing skills. These variables were used to estimate a network and the main centrality indices. The results confirmed relationships between variables and drawing abilities reported in the original work. Moreover, they indicated variables with a significant impact on the network dynamics. Overall, the results showed that the PNA approach helps to identify structures in the data even in the absence of specific assumptions and can be used as a schema to guide targeted interventions focused on the enhancement of target dimensions (e.g., drawing skills).