Reducing ginning energy use through cultivar improvement could reduce ginning and energy cost. The objective of this study was to detect genetic variability for ginning energy and ginning rate. Thirty-four conventional and 12 transgenic genotypes were evaluated in 2008 and 2009 for ginning energy requirements and ginning time rate. The experiments were conducted at two sites near Stoneville, Mississippi. Field plots were one row 12.2 m in length and 1.0 m between rows. Ginning efficiency was based on measurements of ginning energy (Wh kg-1 lint) and ginning rate (g lint s-1). The mean square values for genotypes were significant for gross ginning energy, net ginning energy, ginning rate, and all other traits studied. The two genotypes with least ginning energy were 'AR 9317-26' and 'Yugo 8' with average ginning energy of 7.5 and 7.9 Wh kg-1 lint, respectively. The fastest ginners were 'MD 25' and 'FiberMax 960 B2R' with 3.35 and 3.32 g s-1 lint, respectively. Fuzz percent, fiber strength, fiber length, neps, and fineness were highly correlated with ginning energy. Fuzz percent, fibers seed-1, lint percent, boll weight, and neps were highly correlated with ginning rate. The correlations of fuzz percent with ginning energy, r = 0.62, and ginning rate, r = -0.40, appear to be useful tools in improving overall ginning efficiency.