Particle-phase pressure has to be considered in two-fluid models describing particle-fluid flows, and is important for the stability of such flows. However, most research so far has focused on its dependence on shear-induced particle velocity fluctuations. This Brief Communication proposes a simple model to predict collisional particle-phase pressure at high particle inertia due to microscopic particle number and velocity fluctuations in the absence of any large-scale shearing. The results are in good agreement with the experimental data of Zenit et al. ͓J. Fluid. Mech 353, 261 ͑1997͔͒.Particle-fluid two-phase flows are widely encountered in both natural and industrial processes. Two-fluid models 1,2 are mainstream simulation approaches to such systems, which treat both the particle and fluid phases as interpenetrating continua. Therefore, quantification of the particle-phase stresses generated by particle velocity fluctuations and hence collisions among the particles, becomes a necessity of the establishment of a two-fluid model. It has also been recognized that particle-phase pressure is crucial to the stability of particle-fluid flows. 3 Kinetic theory of granular flows ͑KTGF͒ has been widely used to predict particle-phase stresses, and how to quantify particle fluctuation velocity or granular temperature is the central problem for the theory. However, most research thus far has focused on shear-induced particle velocity fluctuations ͑see Sangani 4 and references therein͒. Theoretical works on particle velocity fluctuations in response to microscopic particle number fluctuations in the absence of largescale shearing have received much less attention, especially at high particle inertia. 5 For particles with small inertia, the characteristic viscous relaxation time is very short compared with the characteristic particle collision time, and the particle fluctuation velocity is dominated by anisotropic and long-range correlated particlefluid interactions, which can be well predicted by Segrè's model. 6 The model has been based on the balance between the rate of gravitational potential increment in a correlated spatial region and the rate of viscous dissipation. At the other extreme, direct interparticle collisions should be the dominant mechanism for random and isotropic particle velocity fluctuation and energy exchange, if the collisions are nearly elastic. 7 Koch 8 analyzed dilute gas-solid flow with ideal interparticle collisions by the kinetic theory, and then extended his theory to dense gas-solid flows at small particle Reynolds numbers ͑Re= g U p d p / g , where g is fluid density, U p is particle velocity, d p is particle diameter, and g is fluid viscosity͒, and finite Stokes numbers ͓St= p U t d p / ͑9 g ͒, where p is particle density, U t is particle terminal velocity͔ with inelastic interparticle collisions. 9 Buyevich and Kapbasov 10 showed that their model can qualitatively reproduce Zenit's experimental data 11 with an adjustable numerical coefficient. However, as Sundaresan concluded in a recent re...