Twenty one oats genotypes were evaluated in a Randomized Complete Block Design to analyse the functional relation and existence of variation among yield and yield components through linear regression and correlation analysis during rabi, 2012 - 2013. In this study, based on the experimental data, ANOVA was worked out for different characters. Different variability parameters viz., mean, range, standard error of difference (SEd) and critical difference (CD) were calculated. The number of groats per panicle exhibited highest mean square value 996.3** than other characters which signified at 0.01 probability level. Among the twenty one genotypes, TNAs 1 recorded highest harvest index (44.1) and single plant yield (34.4 g) than other genotypes included in this study. So TNAs 1 observed was genetically different genotype. Among the characters studied, panicle weight showed the least SEd (0.19) value followed by 1000groat weight (0.72), panicle length (0.77) and single plant yield (0.80), which showed the least influence by environmental variations. Hence, the selection of lengthier panicle need not result in yield increase. So, selection of genotype with heavier panicle will be a fruitful exercise. According to correlation coefficient analysis, single plant yield is positively and significantly correlated with number of productive tillers per plant (r = 0.85**), total number of basal tillers per plant (r = 0.74**), number of groats per panicle (r = 0.589**), dry matter yield per plant (r = 0.481*) and panicle weight (0.34*). These traits should be used as selection criteria to improve oats genotypes with higher grain yield potential. There is a highest linear regression existed between single plant yield and dry matter yield (R2 = 0.35%). Existence of significant linear regression between these pairs of characters exhibited the presence of functional relationship between these characters. These traits should be used for phenotypic selection to improve oats genotypes with higher grain yield potential.