Purpose: Intentional communication acts, or purposefully directed vocalizations and gestures, are particularly difficult for infants at elevated likelihood for eventual diagnosis of autism. The ability to measure and track intentional communication in infancy thus has the potential to aid early identification and intervention efforts. This study assesses the validity of a novel measure of intentional communication intended for use within semistructured caregiver--infant interactions. Method: The Intentional Communication Coding Scale (IC Coding Scale) captures infants' Vocalizations, Gestures, and Combined (vocalizations paired with gestures) acts. Using data from 36 infants at elevated likelihood for autism, we tested the convergent and discriminant validity of the IC Coding Scale with established language measures using Spearman's rho. We tested interrater reliability using intraclass correlation coefficient (ICC) calculations. Finally, we tested responsiveness (i.e., sensitivity to change) using Wilcoxon signed-ranks tests and Spearman's rho. Results: Our initial psychometric tests suggested adequate levels of convergent and discriminant validity. ICCs ranged from 0.77 to 0.92, while confidence intervals were wide, suggesting that Gestures and Combined acts were coded more variably than Vocalizations were among raters. Tests of the scale's responsiveness suggested adequate sensitivity to change across a 12-week period. Conclusions: This study is an important first step toward validating the IC Coding Scale for use in measuring and tracking intentional communication behaviors in infants at elevated likelihood for autism within more naturalistic, semistructured activities. Additional studies are needed to disentangle the effects of intervention from maturation and to examine types of intentional communication acts in more detail.