Previous workers have shown that the net particle flux within the riser of a circulating fluidized bed (CFB) may be measured by a simple technique with a suction probe facing upwards and then downwards. However, this technique does not give accurate measurements of the upward and downward particle fluxes independently. A probe similar to that of previous workers is used here to measure the flux of the upflow. However, the flux of the downflow is measured with a probe through which no external air flow is imposed. The method presented here uses an 8-mm tubular probe, penetrating the riser wall and rising at a sharp angle. The probe is constructed of sections of steel tube and 400 mesh screen. Particles fall into the probe and then into a sealed reservoir for subsequent analysis. The gas flow in the riser passes through the screen sections of the tube, so that the gas flow at the tip of the probe is virtually undisturbed by the presence of the probe. The method is verified in a CFB riser of diameter 115 mm, fluidizing sand of average size 209 mu m, with a solids circulation rate in the range of 10-30 kg/(m(2) s) and air velocity in the range of 3.4-5.4 m/s, corresponding to a nominal solids concentration (based on pressure-drop measurements) of 0.5-2.5%. The technique presented here allows for accurate and independent measurements of both the upward flux and the downward flux, and thus represents an improvement on previous methods. Previous workers have specified that a critical surface where the time-averaged net flux is zero demarcates the dilute core from the dense annulus. However, the total particle flux, defined as the sum of the upward and downward fluxes, is proportional to the time-averaged local particle concentration, and so is indicative of physical phenomena in the riser. Trends for the total flux are presented.