Investigations to study the effect of phosphorus and plant density on the growth, yield and protein content of green gram (Vigna radiata (L) Wilczek) were curried out at the Institute for Agricultural Research farm, Samaru (1111 N ;07,38 N E, 686m above sea level) Zaria during the rainy season of 1993 to 1995. The treatments consisted of five green gram varieties (Galadima, VC 2768A, VC1178A, VC2778A and VC 1973A); three plant densities (200,000; 400,000 and 600,000 plants/ha) and three levels of phosphorus (0.25 and 50kgP/ha) arranged in a split plot design in which varieties and plant densities well assigned to the main plot and phosphorus levels were placed in subplot. The treatments were replicated three times. The results showed that variety Galadima was superior to all other varieties for the growth characters studied, while variety VC 1178A was superior to other varieties in terms of yield character investigated with the exception of numbers of days to 50% flowering, pod length, pod width, number of pods/plant and number of seeds/plant. Variety VC 2778A had higher grain crude protein than other varieties. Increasing plant density from 200,000 to 600,000 plants/ha decreased most of the parameters except LAI, NAR, RGR, pod weight/ha and grain yield/ha which were increased. The grain crude protein content was reduced with increased density up to 400,000 plants/ha. Growth characters as well as yield characters and grain crude protein contents were significantly increased with increasing phosphorus rates from 0 to 50kgP/ha. Correlation analysis showed positive and significant associations between yield and most of the parameters studied. Indirect contribution of most characters was greatest through total dry matters and 1000-grain weight. Plant height and 1000-grain weight make the highest individual contributions to yield; while the combinations of plant height and total dry matter; and number of seed/pod and 1000-grain weight were the highest contributors to yield. Polynomial responses of grain yield to density and phosphorus were linear and guadratic respectively. Using the regression constant, optimum level of phosphorus was found to be 23.43kgP/ha.