Based on multi-factor orthogonally designed field experimental plots, the correlations among yield components as well as their direct and indirect effects on the seed yield of Bromus inermis L. cv. 'Carlton' were investigated. The seed yield parameters fertile tillers/m 2 (Y 1 ), spikelets/fertile tiller (Y 2 ), florets/spikelet (Y 3 ), seed number/spikelet (Y 4 ), seed weight (Y 5 ), and seed yield (Z) were determined by hand in the field for the years 2003-2005. Via ridge regression analysis, a steady algorithmic model of seed yield with its five components was found that could closely estimate the seed yield. The component Y 1 had the largest correlation coefficient with Z, followed by Y 2 . The contributions of the five components to the seed yield in decreasing order are Y 3 . The inter-correlation among the components Y 1 to Y 5 and Z exhibited significance but Y 1 was not correlated with Y 3 . Therefore, direct selection for large Y 1 , Y 2 , and Y 4 would be an effective means of selection for high seed yield in the grass.