Spatial skills can predict mathematics performance, with many researchers investigating how and why these skills are related. However, a literature review on spatial ability revealed a multiplicity of spatial taxonomies and analytical frameworks that lack convergence, presenting a confusing terrain for researchers to navigate. We expose two central challenges: (1) many of the ways spatial ability is defined and subdivided are often not based in well-evidenced theoretical and analytical frameworks, and (2) the sheer variety of spatial assessments. These challenges impede progress in designing spatial skills interventions for improving mathematics thinking based on causal principles, selecting appropriate metrics for documenting change, and analyzing and interpreting student outcome data. We offer solutions by providing a practical guide for navigating and selecting among the various major spatial taxonomies and instruments used in mathematics education research. We also identify current limitations of spatial ability research and suggest future research directions.