The objective of this research is to prove whether or not Concept Sentence Technique can improve the students' ability in writing recount. This research applied pre-experimental research design by using one-group pre-test post-test design. The population of this research is the tenth grade students of SMAN 4 Palu, and the sample is X MIA 3 consisting of 34 students selected by using purposive sampling technique. Then, the researcher uses writing test as the instrument to collect the data. The test is divided into two parts: pretest and posttest. The pre-test is used to measure the students' ability in writing a paragraph of recount text before treatment, and post-test was used to measure the students' ability after the treatment. The data is analyzed statistically in order to know the significant difference of the students' achievement in pre-test and post-test. The result of data analysis shows that the mean score increased significantly from 41.76 in pretest to 71.47 in posttest. By applying 0.05 level of significance and the degree of freedom (df); 33 (df = 34–1=33), the researcher found that the value of t-counted is higher than the value of t-table (11.78> 1.693) It means that the hypothesis is accepted. It suggests that the use of Concept Sentence Technique is effective to improve the ability of the tenth grade students at SMAN 4 Palu in writing recount text.