Research was conducted from 1982 through 1984 on a Shipps clay soil (Udic chromustert) in east-central Texas to determine the density-dependent effects and critical periods of johnsongrass [Sorghum halepense (L.) Pers. # SORHA] competition with cotton (Gossypium hirsutum L.). Two experiments were conducted each year, one to assess cotton yield response to increasing johnsongrass density and the second to establish the johnsongrass-free requirement of cotton and the johnsongrass competition period. A critical minimum density of 2 plants/9.8 m of row was observed. Full-season johnsongrass competition densities exceeding 2 plants/9.8 m of row resulted in seed cotton yield reductions each year. Yield decline was most rapid as density increased from 2 to 8 plants/9.8 m of row. Yield loss averaged 1, 4, 14, 40, 65, and 70% for johnsongrass densities of 1, 2, 4, 8, 16, and 32 plants/9.8 m of row, respectively. Percent yield loss was proportional to yield potential at johnsongrass densities of 4 and 8 plants/9.8 m of row. Three to four weeks of rhizome johnsongrass competition or 6 weeks of seedling johnsongrass competition resulted in significant yield reductions. Maintaining cotton johnsongrass free for 4 weeks did not prevent significant yield reductions.