There are much research in health sector this time. Some research have goal to asses exposure effect to outcome. Researcher often use regresssion. Regression is a convensional multivariate analysis. Conventional multivariable analyses may not always be the ideal method for estimating treatment effects in observational studies. When there are large differences in the distribution of covariates between treatment groups, adjusting for these differences with conventional multivariable techniques may not adequately balance the groups, and the remaining bias may limit valid causal inference. Adjusted use in the end of regression analysis, some of scientis believe that adjustment can be show. Goal of this research is to compare the result of convensional multiariate analysis versus propensity score matching analysis in case study of infant immunizanon using Data ASUI-I KAP2 2003. The comparison will use the same model, that make model according logistic regression. Model which be compare is model without interaction ariabel. Modell will be analised with propensity score matching. Result of research is there is different betwen Odds Ratio from logistic regwession analysis and Odds Ratio from propensity score matching analysis. The value OR from logistic regression is 0,9898, and the value OR from propensity score matching is 0,9656. Propensity score matching succes make matching 574 sample or 68,27%. The differen is not big bacause effect the confounder are not big. For look effect exposure ar risk factor model its better use PSM because it can reduce selection bias.if you want to analysis determinan factor model, where there are much independent vaiiabel its better use logistic regression analysis because the variabel have same position.