In order to determine phenotypic diversity, 40 Rosa damascena accessions were evaluated in a completely randomized block design with 3 replications at the experimental field of Research Institute of Forests and Rangelands, Tehran, Iran. Several traits including flower weight, flower diameter, peduncle length, number of petals, number of stamens, leaf area, leaf dry weight and thorns intensity were recorded and analyzed for two years (2005-2006). Correlation coefficients showed that flower yield correlated positively with number of petals and negatively with peduncle length, leaf area, leaf dry weight and thorn intensity. Because of high heritability obtained for flower weight, number of petals, number of stamens and thorn intensity, selection based on these characters is recommended. Cluster and principal components analyses showed no relationship between genetic divergence and geographical origins, indicating germplasm exchange between different areas of Iran.