Preparation of speech in advance of actual production has consistently been shown to result in greater speech fluency. This observation is important given the impact of speech fluency in social perception; however, it raises questions concerning the nature of the processes by which communicative behaviors are prepared and of the representation of those behaviors in the cognitive system. The current research represents an attempt to address these issues. In Experiment I subjects provided with an abstract problem-solution sequence exhibited less silent pausing during speech than a control group which was not given such a sequence. A second experimental group provided with an abstract solution-problem sequence exhibited less pausing than the control group, but not significantly so. In Experiment II, increasing practice with the solution-problem sequence was found to lead a decreasing linear trend in silent pausing. These findings are discussed in terms of their implications for understanding the nature of production of communicative behavior.