Purpose We examined the English semantic performance of three hundred twenty-seven 7- to 10-year-old Spanish-English bilinguals with (n = 66) and without (n = 261) developmental language disorder (DLD) with varying levels of English experience to classify groups. Method English semantic performance on the Bilingual English-Spanish Assessment-Middle Extension Experimental Test Version (Peña et al., 2008) was evaluated by language experience, language ability, and task type. Items that best identified DLD for children with balanced and high English experience were selected. Separately, items that best identified children with high Spanish experience were selected. Results Typically developing bilingual children performed significantly higher than their peers with DLD across semantic tasks, with differences associated with task type. Classification accuracy was fair when item selection corresponded to balanced or high level of experience in English, but poor for children with high Spanish experience. Selecting items specifically for children with high Spanish experience improved classification accuracy. Conclusions Tailoring semantic items based on children's experience is a promising direction toward organizing items on a continuum of exposure. Here, classification effectively ruled in impairment. Future work to refine semantic items that more accurately represent the continuum of exposure may help rule out language impairment.