To determine potential traits and selection criteria for yield improvement, twenty seven sesame lines and two checks were grown in two successive seasons of 2011 and 2012 to assess the presence of variability for desired traits and amount of variation for different parameters. The two experiments were grown in the Agricultural Production and Research Station of National Research Center, El Nubaria Province, El-Behera Governorate, Egypt, respectively. The data collected on 10 characters were subjected to three different analyses. Variability among the lines were determined with respect to the studied parameters. The results showed significant genetic variances and covariances among lines for all studied traits for all characters except, days to maturity in F6 and fruiting zone length in F7, indicating the existence of variability and the potential for selection and improvement within characters. A considerable amount of variability among lines for the traits studied also indicated the usefulness of selection for these traits in the genetic material used for future improvement in sesame. In variability analysis, high heritability was accompanied by a high genetic advance and a high genotypic coefficient of variation (GCV) for number of branches plant-1, stem height to 1st capsule, number of capsules plant-1 and seed yield plot-1. In correlation analysis, the genotypic correlation coefficient was generally higher than the corresponding phenotypic correlation coefficient which indicated that the apparent association might be due to genetic reason. In selection indices analysis, twelve conventional selection indices (Ii) and two restricted selection indices with different number of traits were evaluated. Selection index was constructed to help selecting for several important traits simultaneously. Judicious use of conventional and restricted selection indices is of a great importance to the breeder. Selection of F6 line (s) that combine high seed yield with one or more desired agronomic or seed yield component trait in a new cultivar was practiced via 12 conventional and two restricted selection indices. These lines were evaluated in the next generation in an evaluation experiment. Heritability values ranged from 23.2 % for days to maturity to 92.2% for stem height to 1st capsule in F6 and from 45.3% for fruiting zone length to 93.9% for oil % in F7. There are no discrepancies between expected (GE) and actual (GA) genetic advances reported for four traits viz., fruiting zone length, branches plant-1, capsules plant-1 and seed yield plot-1 when direct selection for individual traits was applied. Seed yield plot-1 was significantly and positively correlated with plant height, stem height to 1st capsule, branches plant -1, capsules plant-1 and yield plant-1. Contradictory, seed yield plot-1 was significantly and negatively correlated with both seed index and oil%. Besides, the highest relative efficiency (591) was calculated in the conventional selection experiment that contained four traits viz., plant height, branches plant-1, capsules plant-1 and seed yield plot-1. Two restricted indices were applied for the characters that showed negative association with seed yield plot-1. Holding seed index and/or oil content of selected sesame plants to the means of unselected plants in a restricted selection index increased the seed yield plot-1 by 3.0 and 1.3% of lines mean while conventional selection index decreased seed yield plot-1 by 5.0 and 3.0% of lines mean, respectively in the selection experiment. A rank correlation between F6 either for direct selection for seed yield or among all selection indices used (conventional and restricted) and rank of seed yield plot-1 in F7 was positive and significant or highly significant. The line C3.8 recorded the highest seed yield plot-1 and higher resistance for foliar disease amongst the 27 elite lines in F6 and F7. It was higher than the mean of the population of lines by 28.4 and 44% and than the two checks cultivar by 48 and 100 % in 2011 and 2012, respectively