This study is an empirical investigation of both the distribution of the specific direct object marker rā in restrictive relative clauses in spoken Farsi and the degrees of acceptance of different patterns of this structure among 50 Farsi speakers. Examples of the target structure are taken from modern films and a spoken translation test and then used in an acceptability judgement test. The findings show that there is a considerable disparity between the written relative clauses studied in previous research and spoken versions of the same structure. More specifically, the findings in this study show that the omission of the rā marker is common, that the use of rā-marked referent pronouns no longer seems to be an acceptable feature of spoken Farsi, and that the use of the rā marker and the judgement of its acceptability change according to the syntactic function(s) of the head noun phrase.