In this study I investigate the effectiveness of two communicative grammar tasks. I examine the tasks in terms of the language use and I determine the students' preferences regarding them. One of the communicative grammar tasks was designed to elicit the use of prepositions and the other was designed to elicit the use of relative clauses. Two groups of students participated in this study: native speakers from different various fields of study and non-native speakers (from two proficiency levels) enrolled in an Intensive English Orientation Program. Both groups performed the two tasks and the frequencies of the targeted structures were compared to the results of a corpus-based study of a face to face conversation. A comparison of results was also done among the two groups of participants. These comparisons were done in order to identify if the tasks naturally and to a reasonable extent elicit the use of the structures not only among the native but also the non-native participants. The non-native speakers also performed two traditional structure oriented exercises and completed a questionnaire where they expressed their preferences for the traditional activities or the communicative grammar tasks. This was done in order to determine the motivational appeal of the communicative grammar tasks. The results obtained suggest that the two communicative grammar tasks used in this study are effective activities. In general, they elicited the targeted structures and the students chose them over the traditional exercises. However, there are differences among the groups of participants.;On the whole the tasks elicited the structures more with the native speakers, followed by the high level non-native speakers and finally by the intermediate level non-native speakers. As for the students' preferences, the communicative grammar tasks were chosen over the structure-oriented exercises, although the preferences were stronger for the high level students than for the intermediate level students. Further research is necessary, however, in order to confirm the patterns obtained with these participants.