Various levels of skips ranging from over 0.3 to 1.2 m in length were established in 'McNair 235' cotton (Gossypium hirsutum L.) planted April 29, 30, and 23 in a three-year study initiated in 1985. Skips were indexed according to length (skips 0.3-0.45 m were assigned a value of 1 with index value increased by 1 for each additional 0.15 m of skip length). Plots with total index sums of 20, 40, and 60 per 18.3 m of row were compared to plots with optimum stands (no or very few skips) and to plots reworked and replanted two weeks after first planting. Plots consisted of four rows, and skips were established in all four rows. In 1985 and 1987, yields from replanted plots were significantly less than from all originally-planted plots. Yields from replanted plots in 1986 were lower than from all originally-planted plots except those that had skips making up 80-90 percent of the 18.3 m of row length (index sum of 60). In all three years, yields from plots with optimum original stands did not differ from those with a skip index total of 20 but were significantly greater than those with a total index of 60. In 1986 and 1987, yields from plots with an index sum of 40 (60 percent of row missing) were equal to those with optimum stands. Results obtained were similar to those recorded in an earlier study (1980-84) except skips appeared to have less effect when in all four rows of plots than when in two center rows only. Results also suggest that growers should probably consider keeping original stands unless they are drastically reduced and replant only as a last resort.