Abstract The transformation to equivalent dimensions, an innovative methodology to investigate earthquake clustering, is engaged in tracking the preparatory process of the 2017 Mw 8.4 Tehuantepec, Mexico, mainshock. The studied seismic series comprises earthquakes from 1999 to May 2020 in an area extended three times more than the main rupture length. We parameterized each earthquake by its magnitude, the waiting time for its occurrence, that is, the time since the previous earthquake after which this earthquake occurred, and the epicentral offset from the previous earthquake, that is, the distance of the epicenter of this earthquake from the epicenter of the previous earthquake. After transformation to equivalent dimensions, these three parameters became mutually comparable. The average distance between earthquakes in the space of these three transformed parameters was used to determine the degree of event clustering. We calculated this distance in data windows moving in time, each consisting of 100 events. The average distance exhibited a consistent upward trend from ten to two years before the mainshock. Then, it declined until the mainshock. This precursory up–down signal was highly significant statistically. We showed that the detected time changes of the average distance resulted from the evolution of the earthquake clustering in the space of their parameters.