Power consumption, gas holdup and oxygen mass transfer in agitated gas-liquid columns have been studied for an air-water system. Measurements have been carried out in a reciprocating plate reactor using five different types of perforated plates and in a stirred tank reactor with one, two and three Rushton turbines, a helical ribbon impeller with and without surface baffles. Each mixing vessel had an identical geometry with a working volume of 17 L. For reciprocating plate stacks, the gas holdup is a complex function of the perforation diameter, the frequency of agitation and the gas superficial velocity. For radial-type mixing devices, the gas holdup increases more rapidly with the speed of rotation for the helical ribbon. The power imparted to the fluid by the mixing device is independent of the gas superficial velocity for the plate stacks and the helical ribbon impeller for a given frequency or speed of agitation whereas it decreases for Rushton turbines. The correlation of the power consumption obtained for all mixing devices plotted against the reciprocating frequency or speed of rotation to the third power shows a linear fit. K(L)a values were correlated very well with the power input per unit volume and superficial gas velocity for all mixing devices. At lower power input per unit volume, K(L)a is a function of only the gas superficial velocity. At higher input power per unit volume, K(L)a increases rapidly with an increase in the intensity of agitation. Reciprocating plates with larger diameter perforations led to higher K(L)a values whereas the lowest K(L)a were obtained with the helical ribbon impeller. Correlations for one and three Rushton impeller assemblies were almost identical whereas measured K(L)a were much higher for the two-impeller assembly due to the presence of a highly mixed zone in the vicinity of the dissolved oxygen probe.