MAN has conducted research in Bradford 1 aims to find and how much influence the application of Active Learning SAVI based on the results of class X studying chemical redox subject matter. The study population numbered ten in class X class. This experimental method with random sampling cluster sampling, obtained by X-6 class of experiments are given in the class-based learning Active Learning SAVI, and the class of X-8 in a conventional classroom learning is given control. Two kinds of methods of data analysis, the preliminary data analysis including test for normality and homogeneity test results of eight classes normally distributed and homogeneous. Final data analysis also includes the results of normality test normal distribution of data, test the equality of two variances of two groups of results have the same variance of 1.1311. Hypothesis test is obtained by analysis of biserial correlation coefficient of 0.5269, the coefficient of determination of 27.76%, thoroughness studied in the classical style in the experimental class at 92.6%, while the control class by 76%. The average total score of affective and psychomotor aspects, experimental class achieved better grades than the control class. Based on opinion poll of students gained an average of students strongly agreed with the application of Active Learning teaching methods based on students Savi. Therefore, the experimental class students learn better than the control class.