Some simple step-wise procedures are presented for exploring the conformational spaces available to small peptides. For the generation of possible main-chain conformers we review the use of our general-purpose program RAMBLE (and its peptide modelling front-end, PEPSIM), and for the specific task of simulating disulphide-rich peptides we report on a new program, POSSIP. We then describe a detailed investigation of the conformational preferences of the buried amino-acid side-chains in protein crystal structures, and show how the resulting library of favoured rotamers can be employed to speed the job of modelling peptide side-chains.