The purpose of this study is to investigate mapping of oil rose (Rosa damascena) parcels in high-resolution satellite imagery using two different methods. This study was carried out within boundaries of Guneykent Municipality in Isparta in where was produced most of oil rose production in Turkey. Quickbird-2 satellite image was used as base cartographic, and ERDAS and e-Cognition software were used classification of satellite imagery. This purpose; rectify, pan-sharpen and histogram equalization processes were made in the satellite data. The base map including borders of parcel at 1/5000 scale was produced to orient in field work. Then field survey was carried out using these bases. In field survey, Land Use Type (LUT) was determined for each parcel. The digital LUT map was produced using ArcGIS software. Then satellite image was classified to separate the rose parcels. In classification, the methods of pixel and object-based classification were performed on 4, 3 and 2 band combination of Quickbird-2 satellite data. The most appropriate classification method was selected to apply in study. The manufacturer accuracy, user accuracy and kappa value of oil rose class were found respectively as 48.72%, 18.63% and 0.1539 using maximum likelihood decision rule algorithm of supervised classification method. The boundaries of rose parcels were determined best accuracy by using scale: 100, compactness: 0.5, shape: 0.1 parameters in object-based classification. In the thematic maps, accuracy of rose parcels was found 60.78%. In addition, rose rows were separated by using scale: 25, compactness: 0.5, shape: 0.1 parameters. In the pixel-based classification methods, the rose parcels were mixed with other LUT using high-resolution satellite image. These classification methods were not enough successful for determining of rose parcels. But object-based classification methods were found as applicable to identify rose parcels.