Randomized complete block designs (RCBD) have frequently been used in preliminary yield trials of different crops which are conducted at National Agricultural Research Centre (NARC), even though these experiments comprised relatively large number of entries/treatments. RCBD becomes less efficient for large number of treatments to estimate treatment effects compared with income-plete block designs. The objective of this study was to estimate the relative efficiency of an Alpha lattice design compared with a RCBD when evaluating common parameters of groundnut. Preliminary yield trial on groundnut was conducted at NARC. Estimated relative efficiency (ERE) values were calculated for plants/plot, maturity percent, dry pods yield (g), 100-kernel weight (g), shelling percentage and 20-pods length (cm). ERE values (1.10, 1.14, 1.26 and 1.17) for maturity percent, dry pods yield (g), 100-kernel weight (g) and 20-pods length (cm) respectively indicate where the precision to estimate treatment effects is gained significantly from using an alpha lattice design instead of RCBD. Whereas ERE values (0.99 and 0.90) respectively, for the remaining parameters; plants per plot and shelling percentage indicated that the precision between Alpha lattice design and RCBD was similar.