The evaluation of large rock magnetic datasets may benefit from the application of multivariate statistical methods. The combined use of fuzzy c-means clustering (FCM) and non-linear mapping (NLM) have considerable merit over conventional multivariate techniques and appear to provide a more logical choice. An example of the method of identification and characterization of early diagenetic effects on marine sediments is given using samples drilled at Ocean Drilling Program (ODP) Site 904. This confirms the clusters previously identified more laboriously by multi-variable analyses and detects clusters (diagenetic, detrital, etc.) that may only be recognized by the consideration of an array of rock magnetic data. In another example of a complex diagenetic signal the link is expanded between two papers on a long piston core (KC-01B) from the Calabrian Ridge (Central Mediterranean), where FCM and NLM provide a clue as to why the record of the lower Jaramillo Subchron is delayed in this core.