Basic issues of conformational sampling by using the Monte Carlo Metropolis algorithm will be discussed, with the focus on structural descriptions of nucleic acids and how computational efficiency is affected from this. We then will introduce the concept of moving repeating units as rigid bodies and using their internal flexibility for chain closure. Two pseudorotational parameters describing flexible sugar rings add to six rigid body variables, thus resulting in a nucleic acid model with eight independent degrees of freedom per nucleotide. The implemented energy calculations are based on the Amber force field combined with a simplified continuum electrostatic approach including explicit counter ions. The results obtained for double-stranded DNA let us conclude that the algorithm allows for an efficient sampling of the conformational space and could be used as an alternative to computationally much more demanding molecular dynamics simulations for studying conformational fluctuations and transitions in nucleic acids and how they dependent on base sequences.