This study aims to find out the use of cohesive grammatical devices in twelfth-grade students’ explanation texts. In addition, this study also aims to explore the types of grammatical cohesion specifications in their essays. In this study, the researcher wanted to find out how good the quality of their student’s writing explanation texts was. To achieve this goal, the researcher used the conceptual framework of Halliday and Hasan (1976) regarding grammatical cohesion. This study applied to the twelfth grade of the Senior High School Level. The research uses descriptive and analytic studies by assigning students to write the texts. The result from the students’ explanation texts as the data in this research, there are 146 cohesive grammatical items found in students’ explanation texts. The reference occurs 68 times which has 46.5 of percentages. Conjunction occurs 53 times with 36.3% percent. Substitution occurs 23 with 15.7% percent. Ellipsis occurs two times with the lowest rate of 1.3%. The result shows that all four types of grammatical cohesion appear in students’ explanatory texts, which are the primary data in this study. However, there are subtypes of cohesive devices that don’t exist in students’ explanation texts from all of those devices. The most dominant was the reference and conjunction; on the other hand, ellipsis was the smallest presentation among grammatical cohesive, and only a few students used them. The result also indicated that the lack of grammatical cohesion devices used in terms of the generical structure of explanation text, knowledge, and ability in writing leads the college students to use inappropriate grammatical cohesion devices to be applied to this type of text. Thus, this study comes up with feedback to teachers that the discussion on the structural writing of texts and the use of coherence and cohesion should be more intensive.