Location determination in outdoor areas can be obtained accurately through GPS (Global Positioning System) systems. However, GPS systems cannot accurately measure location information in indoor environments. Special network systems are designed for location determination in indoor environments. In this article, the fingerprint of the environment was obtained by means of the sensors located in an indoor environment and the sensor node locations were determined using ANN. In the fingerprint method, the signal strength of each reference point in the indoor environment, also called the fingerprint, is measured and the fingerprints of the reference points are collected beforehand and stored in a database. Then, signals measured from any location in location detection are matched with previously collected fingerprints. Machine learning algorithms are often used for this mapping. In this article, location determination was carried out by means of ANN based fingerprint algorithm according to RSSI values obtained from grid points of 0.5x0.5m2 in a indoor environment of 5x8 m2 . The method was applied for 2 scenarios. In the first scenario, closed environment is empty, and in the second scenario, an environment with various objects and human mobility is used. In the first scenario, there is no object and human mobility in the indoor environment. RSSI value was measured 10 times in each of the 25 randomly determined coordinates. The average values of the RSSI value at any point were calculated and given as an introduction to the ANN model. The distance between the estimated coordinate obtained from the output of the ANN and the actual position is the estimated error. In this scenario, the total error for 25 points is 455 cm and the average error was 18.2 cm. In the second scenario, an environment with office materials and human mobility is used. As in the first scenario, measurements were made from a total of 25 coordinates. The average RSSI measurements were given to the ANN input and the estimated coordinates were determined. Total error 605 cm average error 24.2 cm was obtained.The proposed algorithm and application produce parallel results with current techniques in terms of accuracy and reliability. The method can be improved by taking into consideration suggestions such as optimizing ANN, fingerprint step, choosing smaller environment grids and making more measurements, and location error can be reduced to minimum levels.