There are three nonparametric regression approaches, namely parametric, nonparametric, and semiparametric regression. Nonparametric regression allows the response variable to follow a different curve from one predictor variable to another. In paired data, the components of the predictor variables and response variables are assumed to follow unknown data patterns, so they can be approached with kernel-based regression models and Fourier series. The basic components are approached with kernel functions and Fourier series functions. Errors are assumed to be normally distributed with zero mean and constant variance. The originality of this research is to obtain a mixed kernel and Fourier series model estimator and then apply it to poverty data in Bali Province. The research stage method begins with a nonparametric regression model estimator based on kernel and Fourier series. The next step is to research the regression curve estimation and obtain lemmas and theorems. The results of the function estimation are highly dependent on the bandwidth, smoothing, and oscillation parameters. In the application to the case of real data, the resulting model gives an R2 value of 0.6278, meaning that the variables used can explain the model by 62.78 percent. From the modeling results obtained, the Open Unemployment Rate has a positive effect on the percentage of poor people in Bali.