To improve a complex character such as yield with low heritability, indirect selection using other characters and appropriate selection index based on several characters have been used effectively. These characters must have effective role in yield stability. This study was conducted to evaluate different selection indices and selection methods for 30 bread wheat cultivars. The experiment was conducted in randomize complete block design with three replications at Research Farm of Faculty of Agriculture of Isfahan University of Technology in 2005. The results showed significant differences among cultivars for number of days to heading and maturity, plant height, biological and grain yields, harvest index, chlorophylls a and b, leaf properties and remobilization rate at 1% probability level and for minimum fluorescence at 5% probability level. Therefore, adequate genetic variation was observed for all measured characters; consequently, there were possibility for selection and evaluation of the selection indices. For grain filling rate and other properties related to grain filling period, the logistic equation had the best fit to the data. Results demonstrated that the selection for increase of grain filling rate and effective grain filling period, may result in increase of grain yield. Correlation between the number of spike/m 2 , biologic yield, harvest index, 1000 grains weight, plant height, leaf area index, and remobilization rate with grain yield was positive. According to the selection indices, Alvand and Shah Pasand were selected as the best cultivars. Number of spike/m 2 had a positive and high coefficient in the selection indices. The coefficient of days to heading in selection indices was negative. Accordingly, selection based on indices may result in selection of early maturing cultivars with massive number of spike. The results of direct and correlated response to selection showed that under similar conditions to this experiment, the early maturing cultivars with massive number of spike, less grains per spike and high 1000- grain weight, may have more yield potential. In this survey, Smith-Hazel index had a higher efficiency.