Flow-through columns were used to assess potential long-term trends in 90Sr biogeochemistry and transport in a Finnish near-surface very low-level waste (VLLW) repository concept. Experiments simulated the effects of water intrusion and flow through the repository barrier and backfill materials, examining impacts on 90Sr migration. Artificial rainwater containing 2.0 mg/L stable Sr (as a proxy for 90Sr) was pumped through column systems that had varying compositions from a matrix of rock flour (backfill material), bentonite (backfill/sealing material), and carbon steel (waste encapsulation material), for 295 days. Effluent geochemistry was monitored throughout. Sr retention behaviour in all column systems was broadly similar. Sr removal from influent rainwater was marked (~95% removed) at the beginning of the experiments, and this degree of removal was maintained for 20 days. Thereafter, Sr concentrations in the effluents began to rise, reaching ~2 mg/L by 295 days. Further, 56%–67% of added Sr was retained in the repository materials over the 295-day reaction period. Analysis of the effluents indicated that colloids did not form; as such, Sr output was likely to be aqueous Sr2+. Upon completion of the experiment, solid-associated Sr distribution and speciation in the columns were assessed through column sectioning and post-mortem analyses, which encompassed the following: total acid digests, sequential extractions, and XAS analysis. The total acid digests and sequential extractions showed that Sr was evenly distributed throughout the columns and that the majority (68%–87%) of solid-associated Sr was in the exchangeable fraction (MgCl2). This suggested that a major part of the solid-phase Sr was weakly bound to the column materials via outer-sphere sorption. Interestingly, a smaller amount of Sr (7%–23%) could only be extracted by aqua regia, suggesting that a proportion of Sr may bind more strongly to the barrier materials. XAS analysis of select samples confirmed that the dominant Sr phase was sorbed to the rock flour and bentonite, but not corroded carbon steel. Columns were also subject to remobilisation experiments using artificial rain- and seawater without added Sr. While rainwater remobilised Sr slowly, high-ionic strength seawater remobilised Sr at much higher rates in the systems containing bentonite. Interestingly, Sr was well retained in the rock flour-only system following rain and seawater intrusion. Overall, the results indicate that the column materials provide reactive surfaces for Sr removal should it be released from waste packages; however, the backfill and barrier materials have limited retention capacity, and the dominant sorption interaction is relatively weak. The safety case for the shallow disposal of radioactive waste should consider the possibility of seawater intrusion and that the bentonite-bound Sr was significantly more susceptible to remobilisation following seawater, despite retaining slightly more Sr during sorption experiments.