The author approaches names from an intra- and inter-lingual point of view to obtain a pertinent classification for translators. Then, he studies over 10,000 names translated from English into Spanish since 1925 until the end of the 20th century within all sort of texts, always trying to understand the constraints which have guided the translators' decision process. The result is put forward as a set of conditional norms which have governed and still govern the translators' behavior in the transference of names. The results leave little room for doubt. Far from a priori automatism, translation is a complex activity in which all sort of macro- and micro- textual constraints take part. The idea of "equivalence-cum-identity" is only a starting point which is constantly negotiated to suit the needs of the target audience.