3 experiments were designed to study the role of contiguity in establishing verbal associations. In Experiment I a transfer test followed learning of a paired-associate list which had compound stimuli (word and trigram). Experiment I revealed positive and negative transfer effects with appropriate and inappropriate pairing of the original stimulus components. Heavy stimulus selection was also apparent. Experiments II and III demonstrated that a pair of items in a discrimination task acquire a bidirectional association although instructions require only that they be discriminated. The data of all 3 experiments support the proposition that contiguous appearance of 2 verbal units will result in the development of an association between them while S is learning other associations. (PsycINFO Database Record (c) 2006 APA, all rights reserved)