Intersensory processing of social events (e.g., matching sights and sounds of audiovisual speech) is a critical foundation for language development. Two recently developed protocols, the Multisensory Attention Assessment Protocol (MAAP) and the Intersensory Processing Efficiency Protocol (IPEP), assess individual differences in intersensory processing at a sufficiently fine-grained level for predicting developmental outcomes. Recent research using the MAAP demonstrates 12-month intersensory processing of face-voice synchrony predicts language outcomes at 18- and 24-months, holding traditional predictors (parent language input, SES) constant. Here, we build on these findings testing younger infants using the IPEP, a more comprehensive, fine-grained index of intersensory processing. Using a longitudinal sample of 103 infants, we tested whether intersensory processing (speed, accuracy) of faces and voices at 3- and 6-months predicts language outcomes at 12-, 18-, 24-, and 36-months, holding traditional predictors constant. Results demonstrate intersensory processing of faces and voices at 6-months (but not 3-months) accounted for significant unique variance in language outcomes at 18-, 24-, and 36-months, beyond that of traditional predictors. Findings highlight the importance of intersensory processing of face-voice synchrony as a foundation for language development as early as 6-months and reveal that individual differences assessed by the IPEP predict language outcomes even 2.5-years later.