To study the correlated traits of oil and seed yield in sunflower (Helianthus annuus L.) single cross hybrids, using regression and path analysis, a North Carolina Design I experiment was conducted in a randomized complete block design in three replications in Agricultural and Natural resource Research Station of Khoy. In 2005, six fertility restorer male lines crossed with 18 cytoplasmic male sterility female lines. Each male line crossed with three different female lines to make two sets with nine hybrids in each set. In 2006, Single cross F1 hybrids were planted for studying correlation of traits. Data were collected from agronomic traits including: flowering initiation, seed filling period, maturity, plant height, stem diameter, head diameter, 1000 seed weight, seed numbers per head, grain/achene ratio, biological yield, harvest index, oil percentage, seed and oil yield. Stepwise regression analysis showed that in the final equation, seed numbers per tray and 1000 seed weight traits with the standard coefficients of 1.457 and 0.733 had significant effect (p<0.01) on seed yield as a dependent variable, respectively. In the oil yield regression analysis as a dependent variable, three traits: seed numbers per tray, 1000 seed weight and oil content with the standard coefficients of 1.147, 0.560 and 0.292 had significant effect (p<0.01) on seed yield, respectively and because of these traits, the others were removed from the last equation. Path analysis of oil and seed yield indicated that for maximum yield, final selection must be done through seed numbers per head and by considering optimum amount of this trait and 1000 seed weight.