Children and adults with and without fluency disorders use filler words (e.g., um, uh) in conversational speech, as well as in response to structured questioning (e.g., What is this?). Several explanations for filler word use have emerged. Clark and Fox Tree (2002) have suggested that speakers use these words to signal to the listener that there will be a delay in completing the message. In typically fluent speakers, the use of uh tends to signal a relatively shorter delay, and um tends to signal a longer delay. Thus, according to this account, these filler words serve a specific pragmatic function. Filler word production has not been systematically investigated in individuals who stutter. In particular, the location and duration of these words in people who stutter has not been examined. If differences were observed in location and/or duration of filler words, this could point to a broader pragmatic function of these words for people who stutter, relative to those who are typically fluent. The purpose of this preliminary study is to compare the frequency, duration and location of filler words in children who stutter (CWS) and children who do not stutter (CWNS). Eighteen children in each group, ages 3:0 to 5:11, participated in a conversational language sample and a structured picture labeling test. Filler words were analyzed for frequency, duration, and location for both tasks. Analyses are currently underway. Results will be interpreted within the context of the potential pragmatic functions of filler words across the two populations.