Seven runs were made with the system Penn. Ref. Co. "Supersol" (dispersed) water (continuous) to investigate the effects of pulse frequency, amplitude plate spacing, volume flow ratio, and throughput on backmixing, using the technique of continuous injection of a tracer solution into the packed section. Backmixing was found to be surprisingly insensitive to all the above variables over the range studied with the exception of pulse amplitude and continuous phase flow rate, showing greatest dependence on the later. Backmixing increased with decreasing continuous phase flow rate and increased with pulse amplitude.