Drillings and sample analysis are a part of exploration activities that reguires large cost and very important toward exploration success. Optimum steps are needed in selecting drilling pattern and sampling technique in exploration. Fan drilling pattern is one of drilling pattern that has its own problem in data distribution. Data in the start point of drilling (drift) will have a high grid density, while in the end of drilling the grid density will decrease. In this case, it will be over-representative in some area that has clustered data such as around the drift. Optimum sampling interval is reguired to solve the problem of data distribution. This research is using secondary data of Cu-Au grade of porphyritic deposit: drilling data set is named GRS-37 consists of 11 drillholes. Geostatistical methods used in this research are Ordinary Kriging (OK) and Seguential Gaussian Simulation (SGS). Block units dimension used for estimation and simulation is 15 m X 15 in X 15 m which adapted from the mine bench dimension. Some scenario of sampling interval used for experiment is 1m, 3m, 5m, 10m, and 15 m. The 'optimum' parameter is decided by estimation variance in OK method and variance of local variability in SGS method. The optimum sampling interval based on estimation variance is 15 m from level 3075 m to 3795, 10 m from level 3000 m to 3075 m, and 5 m level below 3000 m. The optimum sampling interval based on variance of local variability is 15 m from level 3150 m to 3795 m, 10 m from level 3000 m to 3150 m, and 5 m for level below 3000 m. Both methods produced the similar optimum sampling interval for Cu-Au grades in fan drilling pattern.