This study investigates null prepositions in wh-questions and passives. The results of a grammaticality judgment and correction task with 67 Japanese learners of English revealed that null prepositions occurred more often in passives than in wh-questions. Previously proposed analyses of null prepositions (e.g., Dekydtspotter et al. 1998, Klein 2001) are unlikely to explain the different acceptance rates of null prepositions between passives and wh-questions as well as null prepositions in passives, because those studies concern themselves only with null prep in wh-related constructions, such as wh-questions and relative clauses, and hence may not generalize to null prep in passives. The present paper proposes that the higher acceptance rate of null prep in passives may result from difficulty in acquiring a native-like system that prohibits prepositions from assigning Case to their complements. That is, to avoid doubly Case-marked NPs, L2 learners may omit prepositions at a higher rate in passives.