Syntactic abilities vary across individuals. Weak syntax is typically ascribed to limited competence (knowledge) or poor performance (processing). However, with many questioning this dichotomy, alternative explanations should be considered. Arguments related to language exposure are insufficient because language-impaired children often have good input. An alternative account, the learning hypothesis, assumes that individual variation in syntactic abilities reflects variation in construction-learning ability. To evaluate this claim, we tested construction learning in 49 5-year-old English-speaking children, targeting two complex constructions rarely attested in child-directed speech, though with no control of prior exposure. The results revealed that there was substantial variation in the children's construction-learning ability, which was strongly associated with their performance on static standardized language assessments (the Test of Reception of Grammar and the Renfrew Action Picture Task) and that nonadjacent open slots were problematic. While our findings supported the learning hypothesis, further research should determine causes of individual variation in syntactic ability.