Research on spatio-temporal visualization is driven by the develop- ment of novel visualization and data aggregation techniques. Yet, only little research is conducted on the systematic evaluation of such visualizations. Evaluation of such technology is often conducted in real-life settings and thus lacks fundamental requirements for laboratory-based replication. Replication requires other researchers to independently conduct their own experiments to verify your re- sults. In this position paper, we discuss the requirements for repli- cation studies of spatio-temporal visualization systems. These re- quirements are often impossible to achieve for highly contextual visualizations such as spatio-temporal visualizations. We argue that reproducibility—allowing other researchers to validate your findings from your data—is a better aim for highly contextual visualiza- tions. We provide a sample workflow to ensure reproducibility for spatio-temporal visualization and discuss its implications.