Laponite—a synthetic nanoclay with (2:1) layer structure—has shown promise for the improvement of granular deposits susceptible to earthquake induced liquefaction, with much of the work performed in this area relying on testing specimens in which the Laponite is introduced in the sand matrix prior to specimen formation. Field implementation will, however, likely occur through permeation of the porous medium by the Laponite in grout form, and it remains unclear to what degree the results from previous studies can be extrapolated to permeated conditions. This paper presents the results of an experimental program on loose (Dr < 30%) Ottawa sand modified through the addition of small quantities of Laponite (1% by mass of the sand) using two specimen preparation procedures: dry-mixing, in which specimens are formed from sand and Laponite mixed together in dry form; and permeation, in which the Laponite is introduced in the sand matrix in dispersion form after specimen formation. Results of resonant column and monotonic undrained triaxial tests illustrate the effects of the two procedures on the properties of the treated sand over a broad range of strains. The different fabrics generated by the two procedures produce distinct behaviors at small to medium strains. The grain-to-grain bonding previously reported for sand-Laponite specimens is observed only in the dry-mixed specimens. The behavior of the permeated specimens with equivalent Laponite content is controlled by the rheology of the Laponite fluid occupying the pore space at the time of shear, carefully characterized in this study through an extensive program of rheological tests. For these specimens, reduced excess pore pressure and increasingly dilative behavior are observed as the pore fluid develops a gel structure and the yield stress of this gel increases. Independent of specimen preparation technique, all specimens reach the same critical state as the clean sand.