Multi-disciplinarity and multi-proxy approaches are necessary to understand the processes in complex earth systems. However, unlimited and uncontrolled multi-proxy-correlations may be risky. A number of case studies illustrate the potential pitfalls when the processes that drive the individual proxies have no common causal significance or dating is not precise enough. Crossing thresholds at different levels and delay times may also be factors that hamper direct correlations of proxies. Multi-proxy analysis of the intrinsic relationships between proxies in a system is the primary task before any correlation should be made.