In this research, application of remote sensing to agriculture especially for crop type determination was investigated using different classification methods. An agricultural field of SanliUrfa lying in the southeast of Turkey between 36 0 49'-37 0 00' north latitudes and 39 0 04'-39 0 13' East longitudes was selected as the pilot region since this city has the highest proportion of agricultural production of Turkey. As an example, 35% of cotton production, 8% of wheat production and 55% of the peanut production are fulfilled in SanliUrfa. Also, major agricultural fields of Turkey such as Ceylanpinar, Akcakale and Koruklu are located in SanliUrfa and these fields are under the control and management of the Ministry of Agriculture entities. Multispectral images obtained from the SPOT 5 satellite acquired on 22-07-2009 and 24-09-2009 were used in this study. SPOT-5 has vegetation sensitive spectral bands and its 10 meter spatial resolution facilitates the detection of agricultural field boundaries. Different classification methods namely pixel-based and object-based were used in this study to identify the boundaries of agricultural fields, determine the areal distribution of the crops. Also, accuracy and efficiency of the pixel based and object based classification techniques were compared and discussed within different spectral and spatial aspects; using kappa statistics and confusion matrix.