Abstarct: Exploration of minerals and significant mining, such as iron ore mines, require its own equipments, and it is one of the most expensive mines that need considerable time for exploration and mining. since we are trying to determine or detect the presence and position of economically useful geological deposits (like iron ore mines) therefore, we should take care which methods we're going to use.This is possible by applying an appropriate method, incorporating and adapting geological and geophysical information. geochemical methods, gravimetric, geoelectric and remote sensing methods can be used for iron exploration, but mostly we prefer the Magnetic method as the main method, due to the magnetic properties of minerals and magnetic susceptibility in exploration of iron ore deposits .Also, the magnetic method is preferred because of the accuracy of the results than other mentioned methods, for exploration of iron ore mines. Therefore, in this thesis, we're using the magnetic data of 8th anomaly of Sirjan Iron Ore and we will use it in inversion. In short, we can say that the purpose of this thesis is to interpret the results of the inversion of the magnetic data of the 8th anomaly of Sirjan Iron Ore , which can indicate the shape and depth of the anomalies and other geological structures and their horizontal extension. To achieve this goal, we use the Li-Oldenburg algorithm and the MAG3D software. by the help of this algorithm, the geometric shape of the Anomalies and the suseptibility boundary between the Anomalies and surrounding area can be identified. First of all , we perform Forward and inverse modeling on Synthetic data with noise, in order to show the effectiveness of the algorithm. After that, we apply the Li-Oldenburg algorithm to the data of the 8th anomaly of Sirjan Iron Ore and we'll interpret the results. In this algorithm, we can also use the boreholes data to match the results with real geological structures. Also, in this algorithm, for more adaptation of results with geological structures, we use constraints such as positivity, smoothing and reference model.