This study of palatal production in Argentina confirms that there are three types of speakers (voicers, devoicers, and variable devoicers) in Buenos Aires. Furthermore, the results indicate that 30 years after Fontanella de Weinberg's (1978) study of palatals, variation between voiced and voiceless allophones of the palatal phoneme is still correlated with age, but no longer with gender. The fact that the youngest generation almost invariably uses voiceless variants now suggests that a sound change from voiced to voiceless is already complete. Though the innovation of voiceless variants may have started with older speakers, younger speakers appear to be responsible for its rapid spread and have generalized it to the point of being the dominant variant in Buenos Aires today.