This research was conducted to investigate the relationships of some agronomic traits with grain yield in bread wheat and determine the changes in these relationships depending on climate conditions. Twenty four winter wheat varieties were used in this study. Eight agronomic traits were measured along with grain yield. Correlation and path analyses were used to investigate the relationships among the traits of interest. Multidimensional Scaling (MDS) was applied for characterization of cultivars based on observed characters. Results of the statistical analysis showed that grain rate per spike and grain number per spike could be used for selection of genotypes for high yielding. It seems that the effect of days to heading on grain yield varied depending on climatologic factors. Earliness emerges as a highly important trait and it should be taken into consideration when choosing appropriate cultivars for trans-regions such as Canakkale. Multi-Dimensional Scaling may be used as an effective method for characterization of the cultivars and provide useful information for agronomists in recommending the best varieties to producers.