A Monte Carlo technique is applied to simulate the structure of concentrated suspensions of hard spherical pal-tides that obey lognormal distribution. With this technique, the random loose packing, with packing density Phi (m), is obtained first, and then the particles in the packing are randomly separated to achieve a specified solid-volume fraction Phi. The simulated structure is evaluated both in the microscale, the neighboring number distribution, and the distribution of gaps between neighboring particles; and in the macroscale, the distribution of the solid-area fr actions on a series of parallel cross sections. Results show that, at the same solid-volume fraction, the increase in the standard deviation of particle diameters leads to the decrease in the mean neighboring number and leads to the increase in the mean gap. The mean relative gap obtained from the simulation is larger than that from theoretical prediction, delta = [(Phi (m)/Phi)(1/3) - 1]. With particles of lognormal distribution, both the gap sizes and the neighboring numbers distributed over broader ranges than that with equal particles. Results also show that with equal particles and particles of lognormal distribution, there is no significant difference between the distributions of the solid-area fractions on the cross sections. The structures obtained in this study are shown to be completely random, homogeneous, and isotropic by statistical tests.