The influence of large-scale density fluctuations on the evolution of small-scale clustering of galaxies is examined. To investigate the fully nonlinear regime we ran N-body simulations. Numerical results show that the evolution of clustering of galaxies strongly depends on the transition scale of the power spectrum. For a two-dimensional power spectrum with n = -1, the wavelength on which density fluctuations go nonlinear is determined by the value of (lambda(t)lambda(delta))1/2, where lambda(t) is the transition scale of the power spectrum, and lambda(delta) is the length scale on which the rms amplitude of the density becomes equal to unity. We analyze the variance of the power spectrum and potential field in different subsamples of the simulations. To investigate the potential field using observational data, the sample size must be at least of the order of the transition scale in the power spectrum.