This research aimed to determine the effect of the Guided Discovery Blended Learning (GDBL) strategy on concept mastery of the human excretory system. The research method was quasi-experimental with a pretest-posttest control group design. The samples consisted of the eleventh-grade students of Senior High school Jakarta in 2019/2020. Class MIPA-2 was the control class taught using Guided Discovery, and class MIPA-4 were the experimental class taught using GDBL. The learning process had been carried out both online and in the classroom. Research data were collected by integrating essay tests, students' responses to learning strategies, and observation sheets of syntax implementation. The data were descriptively analyzed and tested using ANCOVA. The normality test was performed using the Kolmogorov-Smirnov formula, and the homogeneity test was performed using Levene's test. The statistical tests indicated that the data were normally distributed and homogeneous; thus, ANCOVA analysis using SPSS 24 was performed. The data analysis with the ANCOVA test at 0,05 significance level obtained p value = 0.000, which means that H0 was rejected. The research result implied that GDBL affected students' concept mastery of the human excretory system. It cultivated habits to independently rearrange the concept through teachers' guidance, both from sourcebooks and online media. GDBL can be used as an alternative for biology studies.