Despite the increasing recognition of the quantitative importance of Archaea in all marine systems, the protocols for a rapid estimate of Archaeal diversity patterns in deep-sea sediments have been only poorly tested yet. Sediment samples from 11 deep-sea sites (from 79 degrees N to 36 degrees N, at depths comprised from 469 to 5,571 m) were used to compare the performance of two different primer sets (ARCH21f/ARCH958r and ARCH109f/ARCH 915r) and three restriction enzymes (AluI, Rsa I, and HaeIII) for the fingerprinting analysis of Archaeal diversity using terminal-restriction fragment length polymorphisms (T-RFLP). In silico and experimental analyses indicated that different combinations of primer sets and restriction enzymes provided different values of benthic Archaeal ribotype richness and different Archaeal assemblage compositions. The use of the ARCH109f/ARCH 915r primer set in combination with AluI provided the best results (a number of ribotypes up to four folds higher than other combinations), suggesting that this primer set should be used in future studies dealing with the analysis of the patterns of Archaeal diversity in deep-sea sediments. Multivariate multiple regression analysis revealed that, whatever the T-RFLP protocol utilized, latitude and temperature explained most of the variance in benthic Archaeal ribotype richness, while water depth had a negligible role.