Spatial memory is evolutionarily important for many animals. Boundaries can systematically distort spatial memory in line with hierarchical models, while other evidence supports more metric path-integration theories. To test these competing theories, we developed a novel, highly immersive Object Placement Task (OPT) in virtual reality to examine the effect of boundaries on human spatial memory. Our task decomposes spatial memory into 3 components: 1) object-placement errors, 2) distance judgments, and 3) angular judgments (i.e., our single task provides multidimensional information about spatial memory). Thirty participants used the OPT to recall object positions in environments with or without a central boundary. Bayesian analyses showed that distance influenced memory performance, but boundary presence had no significant effect (Bayes factors favored the null models). These findings suggest a set of conditions in which boundaries may not impact memory. We discuss potential modulators of boundary effects and present our open-source task for future research.