This study examined whether speakers avoid lexical ambiguities in a communication task by avoiding ambiguous descriptions when two interpretations are plausible and, furthermore, whether this avoidance is increased after prior experience of lexical ambiguities as a listener. Forty-eight undergraduate participants described target objects form ambiguous and unambiguous displays to a matcher partner. In ambiguous trials the target object (computer mouse) was presented in a display with a foil object (mammal mouse) that caused lexical ambiguity as the display contained two potential referents. In unambiguous trials the foil object was replaced with a filler object (apple). In order to avoid ambiguous descriptions in ambiguous trials participants had to avoid bare homophone responses (mouse) and produce modified homophone responses (computer mouse) when the display contained two potential referents. Participants completed this task after direct experience (having completed a similar task as a matcher), indirect experience (having watched a video of two people completing the task), or no prior experience of the task, using a between-participants design. Responses were coded as bare homophones or modified homophones and analysed using a mixed-model ANOVA. Results revealed that participants produced fewer bare homophones for ambiguous displays than for unambiguous displays, but there was no effect of prior experience.