In order to evaluate 19 advanced lines of bread wheat, an experiment was carried out in a randomized complete block design with three replications at Gorgan Agricultural Research Station for three years, from 2000 to 2003. Each plot consisted of six rows of 6.6 meter with 18cm between the rows. Simple and combined analyses of variance was done and results showed significant difference (p<0.01) for grain yield among genotypes and genotype year interaction. To determine the stable genotypes, stability analysis was done using of parametric methods (environmental variance and coefficient variation) and non-parametric methods Rank. Path coefficient analysis was done in order to investigation of correlation between characteristics with grain yield and determination of their portion with all characteristics according to simple analysis of variance. The results of this study showed significant (p<0.01) difference among genotypes for evaluated traits. Phenotypic and genetic correlation coefficient between grain yield and traits (1000 KW, plant height, and length of spike, number of spike let, and number of grain per spike, grain weight per spike, grain yield and biomass) were close to each other. This indicated that covariance and environmental variance were minor. The genetic correlation was used for path analysis. Correlation coefficient of grain yield with grain weight per spike, number of grain per spike and 1000KW was significant (p<0.01). The results of path analysis showed that direct effect of grain weight per spike was minor. But direct effect of grain number per spike was high. In addition indirect effects of grain number per spike and grain weight per spike by 1000 KW were high. The results of path analysis revealed that using trait number of grain per spike selection for high yield in Golstan province conditions is very important and this may depend on climate and disease conditions in Golestan province. Result of environmental variance showed that genotypes number 10,8,12 and 20 with lowed value for environmental variance had the highest stability. Based on coefficient of variance, genotypes number 10,8,12 and 20 had the highest stability, respectively. Based on non-parametric method of rank, genotypes number 12 andlO with R of 1.3 and 4 respectively lowest variances and standard deviation (Sd) of rank had the highest stability. Results based on analyses of variance and different stability methods, genotype number 12 with good stability and highest yield of 1140Kgha-l (22.4%) than the check cultivar (Tajan) can be recommended for cultivation in Gorgan region. Results suggest that this line suitable for Gorgan.