1. The precision of the lattice design, with and without recovery of inter-block information, as compared to the randomized complete block design was determined for 22 small grains trials. The average of all tests gave an increase of 9% in precision with recovery of interblock information and a loss of 8% when inter-block information was ignored. 2. Four quadrats harvested from 1/60- or 1/80-acre field plots provided, for the most part, reliable estimates of the yield of the entire plot. The precision of the quadrats as measured by the coefficient of variability is essentially the same as that of the field plots. 3. A good agreement was found for most of the varieties tested when grain yields from rod-row plots were compared with those from field plots and quadrats. 4. Calculations based on the 19 field plot trials showed that increasing the number of replications would be more effective than increasing the number of quadrats per plot as a means of increasing precision. 5. The average precision factors calculated for different numbers of quadrats and replicates were essentially the same for the different cereals, especially for oats, spring wheat, and winter wheat.