Rutile, titanite, and ilmenite are the most common Ti-bearing minerals in metamorphic rocks. Experimental constraints have shown that titanite is stable at low-grade metamorphic conditions, rutile at HP, and ilmenite at HT-LP conditions. Yet, petrological evidence suggests that titanite can also be stable at LT-HP. This implies that both titanite and rutile can be used to develop proxies to track HP metamorphism, which can have interesting applications. In this study, we have investigated the natural occurrence and chemistry of Ti-bearing minerals in gabbroic rocks that record different degrees of metamorphism, including LP amphibole-bearing gabbros from the ocean floor (Mid-Atlantic and Indian ridge IODP LEGs) and from an obducted ophiolite (Chenaillet) and HP Alpine metagabbros including blueschist and eclogite facies rocks from the Western Alps and Corsica. We have performed detailed petrography and Raman spectroscopy and analysed major and trace elements mineral chemistry using EPMA and LA-ICP-MS. We found that rutile is stable at low pressure (< 2 kbar) in ocean-floor amphibole-bearing gabbros, lower than experimental constraints had previously suggested. Rutile is also found in eclogitic metagabbros from the Western Alps and can be chemically distinguished from LP rutile. Blueschist metagabbros from the Western Alps and eclogitic metabasalts from Corsica have titanite stable instead of rutile. While the titanite to rutile transition is pressure and temperature dependent, we demonstrate how small variations in bulk-rock Ti/Ca and Ca/Al values within the NCKFMASHTO chemical system may shift their stabilities. High-pressure titanite from these metamafic rocks exhibits La depletion and low La/SmN values in comparison to titanite from amphibolite-facies mafic rocks. La/SmN or Nb together with Yb and V can be used to distinguish HP titanite from titanite formed under other P-T settings. These new systematics can be useful in studies using detrital Ti-bearing minerals to probe the HP metamorphic record through time.