It is well known that the present perfect (PP) in several Romance varieties has encroached on the semantic domain of the simple past or preterite (Harris 1982, Fleischman 1983). In some cases, this process has essentially gone to completion, eg the French passe compose, and the passato pr6ssimo in Northern Italian dialects are both perfect constructions that now carry out many of the functions previously allocated to the simple past forms in these varieties (cf. Squartini and Bertinetto 2000). In the case of Spanish, it has been argued that the PP in Peninsular (European) Spanish is following the same diachronic path as French (Schwenter 1994a, b), ie that the PP in this variety provides evidence of eventually supplanting the Preterite1 form. In contrast, it is generally assumed that this same process is not occurring in Latin America, where in most regions there is very little, if any, overlap between the functional domains of the PP and the Preterite. However, a less well-known fact about the Spanish situation is that the PP and Preterite do actually show considerable functional overlap in several South American (SAm) varieties of Spanish, particularly those spoken in Bolivia, Peru, and northwest Argentina. Some scholars (section 4) have therefore claimed that the SAm and Peninsular cases represent parallel cases of grammatical/semantic change whereby the PP is gradually taking over the semantic space and discourse functions previously allocated to the Preterite. In this paper, we have two principal goals. First, we intend to provide evidence that distinguishes clearly between the Peninsular and SAm situations: while it is true that the PP has encroached on the semantic/functional space of the Preterite in both regions, it has not done so in the same way. Secondly, we will offer a new analysis of the PP in SAm Spanish which can account for both the functional similarities and contrasts it shows with respect to the Preterite. In doing so, our analysis diverges from previous accounts which fail to distinguish the functions of the two forms (eg Alonso and Henriquez Urena 1951) as well as analyses which claim that these functions show no overlap (eg Escobar 1997).