The development of self-healing concrete (SHC) is intensively implemented to improve stability and reduce maintenance costs. However, the experiments require many parameter variations, increasing the cost. Taguchi methods are used in experimental design to obtain variations of the optimal parameters and reduce the cost of the experiment. The stages of this study are secondary data collection of the increase of compressive strength, testing Taguchi methods, determining contribution parameters using ANOVA, and developing the estimated model. The parameters analyzed consist of the type, the concentration, the application of bacteria, and the curing time conducted as experimental analysis of 2, 3, and 4 parameters. The results of the parameter analysis using the Taguchi have optimal variation results in a row Bacillus subtilis, 105 cells /ml, encapsulation methods, and 28 days. Bacterial concentrations have a very significant value with contributions of 82.94%, 78.90%, and 74.36% in the analysis of 2, 3, and 4 parameters, respectively. The slightest error value was obtained in the analysis of the 3-parameter experiment with a value of 2.99%. SHC modelling obtained two equations consisting of direct mixing and encapsulation, which have a coefficient of determination of 0.89 and 0.82.