The statistical properties of the neoclassical radial diffusion are confirmedthrough direct comparison with a Wiener process by the numerical evaluationsof the cumulant, diffusion and autocorrelation coefficients. Within theneoclassical framework the origin of stochasticity exists only in velocityspace. It is characterized by the stationary, subdiffusive, uniform, and Markovprocesses. Through the drift motion of particle guiding centres, thestochasticity in velocity space leads to that in configuration space, i.e. the radial diffusion. It is shown that such a radial diffusion develops as anapproximately Wiener process, i.e. the statistically non-stationary, normaldiffusive, Gaussian, and Markov process in the asymptotic time region.