Message to CUNY attendees: Hi guys, thank you for your interest in our poster! I (= Katja) will be available for your comments during Poster Session A (Thursday 03/19, 12-2pm EDT, 5pm-7pm CET) via Zoom under https://zoom.us/j/463035435. Short Abstract: There is a clear lack of behavioral evidence on linguistic prediction. Van Berkum et al. (2005) found that younger Dutch adults showed slowing when reading prediction-inconsistent gender-marked adjectives. Here, we used the same paradigm as Van Berkum et al. (2005) to investigate whether younger and older readers of German show processing difficulties when reading gender-marked articles that are inconsistent with the gender of the noun anticipated through discourse. We found that only older adults showed early effects of having predictions disconfirmed. In younger adults, no prediction effect emerged, despite sufficient power to find an effect in the size reported by Van Berkum et al. (2005). Our findings cast doubt on the ubiquity of linguistic prediction and suggest that older adults may be more likely to use prediction during language processing.