We study the clustering properties of the recently compiled Sloan Digital Sky Survey (SDSS) cluster catalogue using the two-point correlation function in redshift space. We divide the total SDSS sample into two richness subsamples, roughly corresponding to Abell R greater than or equal to 0 and Automated Plate Measuring (APM) clusters, respectively. If the two-point correlations are modelled as a power law, xi(r) = (r(0)/r)(gamma), then the best-fitting parameters for the two subsamples are r(0) = 20.7(-3.8)(+4.0) h(-1) Mpc with gamma = 1.6(-0.4)(+0.4) and r(0) = 9.7(-1.2)(+1.2) with gamma = 2.0(-0.5)(+0.7) h(-1) Mpc, respectively. Our results are consistent with the dependence of cluster richness on the cluster correlation length. Finally, comparing the SDSS cluster correlation function with predictions from three flat cosmological models (Omega(m) = 0.3) with dark energy (quintessence), we estimate the cluster redshift-space distortion parameter beta similar or equal to Omega(m)(0.6)/b(0) and the cluster bias at the present time. For the Lambda cold dark matter case we find beta = 0.2(-0.16)(+0.029), which is in agreement with the results based on large-scale cluster motions.