In order to estimate heritability and relationships among traits, 43 diploid sugar beet half-sib families were evaluated in a randomised complete block design with three replications in two locations (Karaj and Miandoab) in drought stress condition. Results of combined analysis of variance showed that there was significant difference between the two locations in terms of root yield, sugar yield, white sugar yield, alkalinity, chlorophyll content, water use efficiency, specific leaf weight and leaf succulence index. There were also significant differences among genotypes for all traits except potassium content, white sugar yield and specific leaf weight. The genotype by environment interaction was significant only for sodium, potassium, alkalinity and extraction coefficient of sugar. The heritability values estimated for root yield, sugar content, white sugar content, sugar yield, extraction coefficient of sugar, white sugar yield and chlorophyll content were 60.61%, 87.10%, 90.88%, 47.90%, 83.91%, 64.0% and 81.11%, respectively. The highest genetic gain belonged to root yield, extraction coefficient of sugar, chlorophyll content and sugar content, respectively. White sugar yield showed a significant positive correlation with root yield, sugar content, sugar yield, extraction coefficient of sugar, white sugar content, chlorophyll content, water use efficiency and specific leaf weight and significant negative correlation with sodium, potassium, alkalinity and molasses sugar. Furthermore, sugar yield, molasses sugar and white sugar content explained 99% of white sugar yield variation and were identified as the most effective traits on white sugar yield. Finally, based on the results of cluster analysis, 43 sugar beet genotypes were divided into three groups with the superior genotypes being placed in the first cluster.