The reconstruction of past human diet is one of the most basic concerns of the archaeologist. Food production and distribution, besides being essential to survival, are defined by, and help to define, a broad range of social, cultural, and political processes. In almost every society, food forms a key part of social and economic relations at multiple levels, as well as serving as a symbol related to different aspects of culture (e.g., ethnicity, political power, religion) and as an expression of forms of social identity (for a review on this topic see, for example, Ashley et al. 2004; Farb and Armelagos 1983; Harris 1985; Mintz and Du Bois 2002). In particular, the use and control of food can become an arena where power and prestige are negotiated (Appadurai 1981; Danforth 1999; Hastorf and Johannessen 1993; Ross 1987). Thus, it is important to recognize that food is a dynamic entity in a wide range of social relations in both private and public spheres. Within the cultural historic framework of Caribbean archaeology, however, most studies have treated diet, subsistence, and food only as fossil indexes of chronological or cultural periods. Rainey (1935, 1940) was the first to do so when he used diet as a marker for shifts in ethnic identity in Puerto Rico. During his excavation in Ponce, Rainey noted that Cedrosan Saladoid subseries assemblages contained high concentrations of crab claws, while those from the later Elenan and Chican Ostionoid subseries contained high quantities of shellfish instead. This led him to define these deposits as belonging to successive peoples of the "Crab" and "Shell" cultures, respectively. In more recent years, deFrance (1989) conducted a more detailed analysis of Cedrosan Saladoid and Elenan Ostionoid assemblages from Maisabel in order to investigate and test Rainey's dichotomy of Crab and Shell cultures. In the 1980s and 1990s, Yvonne Narganes Storde (1985, 1993a, 1993b) also began to conduct more refined zooarchaeological studies intended to identify differences in the diet of distinct cultural groups (Huecoid and Cedrosan) in the early Ceramic age. In all of these cases as well, food remains were used as a diagnostic to define the different assemblages of different, but in this case contemporary, "peoples." While valuable, these studies have failed to view food as an active element in the structuring of interpersonal relationships of the society in which it was being consumed. This reduces food to little more than a passive marker of people, rather than a potentially useful tool for the reconstruction of past social processes. In order to distill meaningful and accurate insights about the broader social dimensions of food in any given past society, one must first employ an analytical methodology that is capable of reconstructing past subsistence patterns with sufficient fidelity and at a satisfactory resolution. Traditionally, paleodietary reconstruction has involved either a reliance on possibly inaccurate or slanted historical sources or the indirect archaeological study of diet through the analysis of food remains (zooarchaeology, paleoethnobotany) or nutrition-related pathologies in human skeletal remains (paleopathology). While these techniques are all capable of providing valuable dietary insights, all also possess theoretical and/or methodological shortcomings that can result in overly coarsely grained and/or misleading reconstructions of past human diet. The study of food remains (either faunal or floral) is, perhaps, the most common means of paleodietary analysis. When carried out diligently, such techniques can provide detailed information on the different types of food resources exploited, quite often with species-specific resolution. This resolution is extremely useful since habitats and possible methods of exploitation can be inferred from the identification of particular species, and when combined with other types of information, such analyses can provide insights into matters such as the control of labor. However, the use of techniques involving the indirect study of diet through food remains can produce nonrandomly biased or misleading results because of the differential survival and preservation of different classes of food (Pearsall 2000). In the Caribbean, as in the neotropics in general, where many important foodstuffs (e.g., manioc) are composed mainly of soft plant tissues that survive poorly or not at all, this is of particular concern (Newsom 1993). As floral and faunal evidence is rarely recovered in the relative quantities in which it was consumed, the conclusions derived from such evidence are necessarily qualitative rather than quantitative (Ambrose 1993:59). Moreover, it is important to remember that the scale of the data derived from the study of food remains is at the level of a community or group rather than that of an individual, that the temporal resolution provided by such evidence can be quite variable, and that the data it can produce can be misleading due to many cultural and natural formation processes that may have. affected its evidentiary basis (Barberena and Borrero 2005). Thus, indirect means of dietary reconstruction will tend to fail in the detection of subtle differences in diet between individuals. Put differently, the study of the actual food remains provides us with the menu rather than the meal (Ambrose 1993:82, 85). The study of health indexes in human bones is another means of acquiring information about past peoples' diets. Since many aspects of health are related to nutrition, skeletal health indicators and pathologies can provide indirect information about the types and quantities of foods consumed. Prevalence rates for skeletal pathologies such as cribra orbitalia, porotic hyperostosis, and linear enamel hypoplasias; the overall disease load in a population; and broader indicators such as stature have all been argued to serve as effective proxies for nutritional well being (Buikstra 1992; Goodman et al. 1988; Roberts and Manchester 1999). While absolute statements about nutrition are not possible as a result of this type of analysis, relative statements based on changes through time or differences between individuals resulting from differences in status are possible. The inferences derived from such studies are necessarily broad, however, and can be fraught with potential difficulties given the multiple etiologies of the pathologies under discussion and the fact that the individuals under study may not be representative of the population at large (Ambrose 1993; Wood et al. 1992). Furthermore, such analysis can really only be used to infer nutritional inadequacy rather than the presence or relative contribution of any actual dietary items (Schwarcz and Schoeninger 1991:284). In light of the methodological shortcomings of the techniques most often employed in the study of diet in the prehistoric Caribbean, this chapter presents the results of a pilot study undertaken to assess the feasibility of instead using biogeochemical analytical techniques to reconstruct the subsistence patterns of the prehistoric inhabitants of Puerto Rico. The analysis of the stable isotopes of two light elements (carbon and nitrogen) in both the organic and mineral phases of human bone was employed to facilitate the direct dietary reconstruction of four individuals interred within the Ceremonial Center of Tibes. This work forms part of a larger ongoing project employing stable isotope analysis of Puerto Rican skeletal materials of varying ages and depositional histories that aims to define the relationship between subsistence and the in situ development of social complexity in prehistoric Puerto Rico. It is hoped that the use of this stable isotope technique, when combined with multisource mixture modeling, will produce data of sufficient resolution and quality to allow for the reconstruction of the possible dietary dimensions of fine-grained social processes. If successful, it is hoped that the application of this technique will allow for the detection of subtle, individual- level changes in diet possibly reflective of a diachronic shift in sociocultural organization.