A two-color Laser Doppler Anemometer (LDA) was used to obtain axial and tangential velocity information in a 0.2 MW pulverized coal flame. In addition to the reacting flow data, a study on the accuracy of using coal as a seed particle to measure gas phase velocity using LDA was performed. Non reacting flow velocity measurements were also obtained near the fuel inlet and in the quart region of a geometrically identical burner to identify the velocity profile at several burner settings and to assist in establishing modeling inlet conditions. Both the reacting and non reacting velocity data were obtained at three or more swirl settings and various axial positions allowing a study of the affect of swirl on inlet turbulence and flame structure. The velocity results were compared with effluent NOx measurements. At the flow rates and accelerations experienced in this study, the coal particles were shown to be useful as seed particles for LDA gas phase velocity measurements. The coal-name Velocity indicated a centerline flame at 0 swirl transitioning to a radially directed flame with a central recirculation zone at swirl settings of 0.5 and 1.5. The transition of the flame structure to a central recirculation zone was also seen at the fuel inlet plane in the non reacting flow studies and was found to correlate with a decrease in measured effluent NOx. Measured axial velocity profiles 5 mm below the fuel inlet showed negetive axial velocities tin the opposite direction of the average flow velocity) were produced along the primary tube as swirl was increased from 0 to 1.5 with the transition occuring between 0.5 and 0.75 swirl. Transition in the now near the fuel inlet correlated well with a drop in effluent NOx and with transitions in the recirculation zones measured further downstream. The strong interaction with burner velocity profiles and NOx suggest velocity profile, in addition to swirl number, is an important measured boundary condition for modeling. The velocity data shown here in combination with a companion paper showing temperature and species data should provide important information needed to develop better models of pulverized coal combustion.