Recipient design theory emphasizes the use of language structure depending on the addressee’s need. We use more sophisticated language when communicating with adults (Ravid, 2005). On the other hand, when there is a child addressee we use less complex language to maintain comprehension (Iverson et al., 1999). Studies also show that different addressees elicit differences in speech and co-speech gestures (Dunst et al., 2012). Findings show that adults use more iconic gestures addressing children compared to adults (Campisi & Özyürek, 2013). Present study investigates how speech and co-speech gestures change in adult-directed speech (ADS) vs. child-directed speech (CDS) for three different tasks. Forty-one participants (Mage=22.04) attended online zoom for two different conditions (ADS-CDS) with three different tasks (story-reading-task, storytelling-task, and address-description-task) in a within design. They were instructed to pretend either communicating to a child or an adult. We hypothesized that participants will use more complex language and less iconic gestures in the ADS condition compared to the CDS. Results showed that, for CDS, participants used more iconic gestures in the story-reading-task (M=.03, SD=.045) compared to ADS (M=.01, SD=.037) (t(40)=-2.618, p=.012). There were no differences for story-telling and address-description tasks. Additionally, language complexity did not differ across tasks and conditions. Findings indicate that speakers might modify gestures considering the addressee depending on task at hand. The difference occurred only for story-reading-task which required the least cognitive load among the three tasks. Results will be discussed at the intersection of language production theory (Bock & Levelt, 1994) and recipient design theory.