This experiment was conducted at the Research Institute of Seed and Plant Improvement, of Karaj, during 1993 and 1994 growing seasons in order to determine the correlations between inbred line traits and their hybrid performance in maize (Zea mays L.).In this experiment, fourty selected inbred lines as female parents were pollinated with two selected inbred line testers as male parents in two isolated fields, and hybrid seed was harvested separately from each field in 1993. Group balanced block design with three replications was used to evaluate the experimental materials. Groups were consisted of Group A (40 inbred lines), Group B (41) hybrids with K1 as a common tester), and Group C (40 hybrids with C103/8 as a common tester). These groups were grown in 1994.The analysis of variance for measured traits showed significant differences between inbred lines and their testcrosses for all traits. These differences were probably due to heterosis in hybrids compared with parental lines (5). The days to silking was longer in inbred lines compared to both hybrid groups, and this delay in maturity was attributed to inbreeding depression (6). The plant height, ear length, leaf width and surface, relative growth rate, ear length and diameter, kernal-row number, and kernal depth had positive and significant correlations with mean.yields of inbred lines, group B, and Group C hybrids. Furthermore, the yield of inbred lines had positve and significant correlations with leaf lenght, number of ears per plant, and 1000-kernal weight but, negative correlation with days to silking. Also, there were positive significant correlations between yield of hybrids in group B with stem diameter and mean yield of hybrids in group C with leafes numbers and lOOWkernal weight.The multiple adjusted coefficient of correlation was significant (R = 0.779 ) between inbred line traits and their mean yields. Because of high and significant R, selection for inbred line traits is necessary before testing for their combining ability. The multiple correlation (R=0.241 ns) between inbred line traits and hyrid performance was low a nd nonsignificant. Therefore, it is concluded that the yield of testcrosses can not be estimated on the basis of morphological traits of inbred line parents.
)