The research consists of two variables, namely the independent variable (Instagram) and the dependent variable (writing ability). The aim of this research is to find out whether or not the use of Instagram is effective in teaching writing Descriptive Text. The population of this research was all grade ten students at SMAN 3 Bekasi, consisted of 370 students divided into 10 classes. The sample of this research was 74 students divided into 2 classes, taken from the population using Cluster Random Sampling Technique. This research used the Quasi Experimental Method, and the design was randomized post-test only control group design. To collect the data, the writers used instrument. The instrument consists of 10 question items in an essay. They were developed from the indicators of writing as specified in BluePrint. To get a good instrument, it was tried out to other classes outside the sample classes, and the instrument was analyzed in the item of its validity, reliability, difficulty index, and discrimination power by using Ms. Excel 2010. After it has been analyzed, it gained 5 items which were used to collect data for the research. The writers used a T-test to find out whether the alternative hypothesis (Ha) was accepted or rejected. The result of this research presented that there was a significant difference of Instagram on students’ writing ability at SMAN 3 Bekasi. The significance (Sig.) score of T-test 0.000 is lower than (<) 0.05. Therefore, the research had a significant difference in Instagram on Grade Ten Students’ Writing Ability at SMAN 3 Bekasi. In addition, Instagram helped the students to develop their writing, and also, it can ease the teaching and learning process.