This research employed a quasi-experimental design and a quantitative approach. The research population consists of 252 students in class VIII from SMP N 2 Karangpucung, who were organized into eight classes. The research sample was chosen using random sampling. The experimental class chosen was VIII.1, while the control class selected was VIII.2 . This research collected data using a pre-test and post-test in speaking direct practice. IBM SPSS V26 for Windows was used to analyze the data. Based on the average calculated N-gain for the control class, it was 09.12 in the effective enough category. Then, the average N-gain for the experimental class is higher than control class, it was 74.66, which in the effective enough category. The hypothesis test using a t-test on both experimental and control classes suggests that Sig (2-tailed) 0,000 < 0,050, suggesting an effect between before and after the treatment of both classes (experimental and control). Even though both treatments are conventional methods and used the Blindfold game as the teaching technique, and it was indeed effective. But the level of effectiveness are different. Therefore, the use of the Blindfold game affects students' speaking skill in class VIII SMP N 2 Karangpucung effectively enough.