Many cotton breeding programs in the U.S. have installed an automatic weighing system in a cotton plot picker to increase operating efficiency in recent years. However, no experimental data are available to document the reliability of such a system. The objective of this study was to evaluate the accuracy, precision, and harvesting efficiency of a two-row cotton plot picker installed with an automatic weighing system with two scales based on results from replicated field tests from 2013 to 2016. Three tests each year, each containing 32 genotypes were arranged in a randomized complete block design in two-row plots by 10 m in length. The 2013 to 2015 results showed a highly significant positive correlation in seedcotton weights between the two rows of the same plot for each genotype harvested and weighed by the two scales in the picker, indicating that the two scales are consistent and reliable. In three tests in 2016, one row of each two-row plot was harvested by the picker, and seedcotton weight was compared with another row harvested by hand. A highly significant positive correlation was detected between the two harvesting methods that had similar coefficients of variation (16.14% for hand harvest vs. 16.90% for mechanical harvest). The two-row plot picker harvested a total of 368 single-row plots (10-m long) in six hours daily, whereas hand harvest by one person averaged two plots in four hours. An average of 417 kg ha-1 was lost due to the mechanical harvest.