The subgrid-scale effects on particle motion were investigated in forced isotropic turbulence by DNS and prior-LES methods. In the DNS field, the importance of Kolmogorov scaling to preferential accumulation was validated by comparing the radial distribution functions under various particle Stokes numbers. The prior-LES fields were generated by filtering the DNS data. The subgrid-scale Stokes number (StSGS) is a useful tool for determining the effects of subgrid-scale eddies on particle motion. The subgrid-scale eddies tend to accumulate particles with StSGS<1 and disperse particles with 11 while has an adverse impact on the particles with StSGS<1. The results show that the Langevin-type stochastic model tends to smooth the particle distribution in the isotropic turbulence.