In five structural-priming experiments, we investigated lexical boost effects in the production of ditransitive sentences. Although the residual activation model of Pickering and Branigan (1998) suggests that a lexical boost should only occur with the repetition of a syntactic licensing head in ditransitive prepositional object (PO)/double object (DO) structures, Scheepers, Raffray, and Myachykov (2017) recently found that it also occurs with the repetition of nouns that are not syntactic heads. We manipulated the repetition of the subject (Experiments 1-3), and the verb phrase (VP) internal arguments (i.e., either theme or recipient, Experiments 4-5) in PO/DO structures. In Experiment 2, the verb was also repeated between prime and target, while in the other experiments it was not. Three different tasks for eliciting the target were employed: picture description via the oral completion of a sentence fragment (Experiments 1-2, and 4), oral completion of a sentence fragment with no visual context (Experiment 3), and oral production of a sentence from a given array of words and no visual context (Experiment 5). Priming occurred in all experiments and was stronger when the verb was repeated (Experiment 2) than when it was not (Experiment 1). However, none of the experiments showed evidence that priming was stronger when either the subject or one of the VP-internal arguments were repeated. These findings support the view that structural information is associated with syntactic heads (i.e., the verb), but not with nonheads such as the subject noun and the VP-internal arguments (Pickering & Branigan, 1998).