This study investigates the effects of Processing Instruction (PI) on the acquisition of grammatical gender and gender-marked pronouns in German. PI was compared to Traditional Instruction, i.e. a traditional, vocabulary-oriented approach using color cues (TI) and a Categorization and Memorization task (CM). The results of an immediate posttest showed that the PI group outperformed both TI and CM with respect to gender assignment on both a gender selection task and a writing task. The PI group also scored higher and responded faster than the TI and CM groups on a comprehension task that required accurate processing of gender-marked pronouns. However, differences between the three groups were not sustained on delayed posttests. These results extend the findings of previous research on PI (e.g. Benati, 2004) by showing that PI can be applied to target forms that are low in communicative value and must be learned item-by-item like grammatical gender in German. Results also lend support to psycholinguistic research that suggests that second language (L2) learners have difficulty acquiring grammatical gender because they do not process nouns together with gender information (Arnon & Ramscar, 2012).