Non-adjacent dependencies characterize numerous features of English syntax, including certain verb tense structures and subject-verb agreement. This study utilized an artificial language paradigm to examine the contribution of item variability to the learning of these types of dependencies. Adult subjects with and without language-based learning disabilities listened to strings of three non-words for which the first and third elements had a dependent relationship. In the low variability condition, 12 non-words occurred in the middle position, and in the high variability condition, 24 non-words occurred in this position. Non-disabled adults were able to learn the non-adjacent contingencies and generalize the underlying structure to new strings, but only when variability was high. Adults with language-based learning disabilities did not perform above chance levels under either variability condition. Thus, this group showed poor sensitivity to statistical information in speech input that both infants and non-disabled adults are known to track. Learning outcomes: As a result of this activity, the reader will: (1) understand the advantages of using an artificial language to investigate language learning; (2) become familiar with a paradigm for studying the rapid learning of syntactic contingencies; (3) comprehend how the ability to map language structure differs for non-disabled adults and adults with a history of language/learning disability as a function of variability in the input the listener receives.