Writing ability is one of the demands of learning English. However, writing is also the most difficult skill faced by high school students in the city of Pekanbaru. For this reason, the learning model should be appropriate in accordance with the age development and problems faced by students. Thus, the authors attempted to analyze; (1) the effect of blended learning on students’ writing ability; (2) the effectof language attitude on students’ writing ability; and (3) the effect of blended learning model and language attitude on students’ writing ability. Referring to the problems and purposes of this study, the author used a quantitative approach namely the experimental method with factorial design. The research was conducted at SMAN 6 Pekanbaru by using a random sampling technique. To measure students' writing ability, the researchersemployed a writing test. Whereas to obtain information or collect data related to language attitude in this study, the researchers distributed questionnaires that have been tested for validity and reliability. After collecting data, researchers tested the data and hypotheses using SPSS 21. In this study, data was normally distributed and the population was homogeneous so that hypothesis testing could be continued. Based on the results of hypothesis testing, this study could be concluded that; (1) there was an effect of blended learning on students’ writing ability; (2) there was an effectof language attitude on students’ writing ability; and (3) there was no effect of blended learning model and language attitude on students’ writing ability.