An experiment to determine the plant characters of wheat that caused differences in yield between certain crop sequences and tillage methods was conducted, holding all other factors constant within season. The plots were planted to the same variety, on the same date, at the same rate (with one exception), and with the same type of drill. The final yield of winter wheat for the years and treatments studied was shown to be closely determined by the number of kernels per unit area. The kernel weight was never found sufficiently low to upset this foremost consideration. These results would not be expected to hold for excessive shriveling or for varieties of varying kernel size. The number of heads per unit area as an individual indicator of yield was exceeded only by the number of kernels per unit area and the number of spikelets per unit area. Variations in number of heads per unit area were much greater than those in the number of spikelets per head and were chiefly responsible for the differences in number of spikelets per unit area. A number of tillers sufficient to produce a large number of heads per unit area is the first requirement for a good yield. The potential number of heads per unit area is indicated from the tillering period but this may be modified by later conditions. Plants per unit area do not usually play an important part in determining number of heads per unit area because they are compensated by the number of heads per plant. The head characters, including spikelets per head, kernels per spikelet, kernels per head, and head length, did not often vary enough to compensate for material differences in head number. In some cases however, the yield level indicated by the number of heads and number of spikelets per unit area was considerably modified by head characters. Conditions seldom remain favorable throughout the entire season. Drought during early growth causes an elimination of weak plants and less vigorous tillers that would otherwise produce heads. The tillers remaining are usually those which will produce heads containing an above-average number of spikelets. A greater number of spikelets per head was observed for treatments and years in which a deficiency of soil moisture was experienced. During the heading stage some yield adjustment in favorable years is made through the number of kernels per spikelet. In comparison with the influence of head numbers on yield such adjustment is very limited. Severe conditions during the heading and flowering period, however, may in some cases cause a reduction in the number of kernels per spikelet large enough to more than compensate for differences in head numbers between treatments. The final adjustment in yield to environmental conditions is made through kernel weight as the plant nears maturity. Significant treatment differences in this character were not apparent, indicating that hazards which reduce kernel weight on one treatment cause similar reductions on other treatments. Seasonal differences in kernel weight were apparent. There appears to be no simple answer to the problem of character response to different methods of cultivation and crop sequence. Detailed plant descriptions at maturity demonstrate that there are many possibilities for plant characters to reflect the moisture and fertility conditions occurring at different stages of plant growth The examination of characters pointed out that continuing adjustments of the wheat plant tend to maintain it in a condition that will enable it to take advantage of the moisture and fertility conditions then existing. For the conditions covered by this study, the number of kernels per unit area, which is the product of the two characters number of heads per unit area and number of kernels per head, was found most practical for estimating yields. When kernel numbers were not considered, the number of heads per unit area and plant height provided the most useful estimate.