Abstract Reported are the results of a multivariate analysis, here called multivariable analysis, of an Italian Late Neolithic archaeofauna. The technique adopted is that proposed by Behrensmeyer (1991), adapted to the needs of zooarchaeology. It is tested on an inter-hut bone accumulation from the target site of Offida, near Ascoli Piceno, Marche region, in central eastern Italy, which is characterized by vertebrate remains, invertebrates, stone implements, and ceramics. The typology of the ceramics date the site from 3800 to 3100 B.C. Offida's subset was chosen not only for its rich and varied information, but also for its containing both domestic and wild mammalian remains. The multivariable method is tested on these two faunal components both separately and jointly, with the addition of the undetermined specimens. The results are then compared with those obtained plotting the variables of the bone accumulations of two open-air Neolithic settlements, a small, temporary one in Italy, Mulino Sant'Antonio, and a large one in Israel, Hefzibah 7–18. The first is an Italian Neolithic site typologically and dimensionally similar to Offida; the second is outside Italy and much larger than Offida. Offida's domestic assemblage results more fragmented than its wild animal component. It shows evidence of intense elaboration by the villagers, low richness and a relatively higher amount of sheep/goats and cattle, no bone articulation or sorting, as well as the effects of moderate exposure and prolonged trampling. The wild animal component, in contrast, shows rather high equitability and a lower overall body size. The analysis proves that wildlife already was a minor resource for the Late Neolithic people in this part of Italy. The multivariable approach also shows that a backyard or pathway assemblage such as Offida's is more similar to the overall bone accumulation of a small-sized, temporary settlement than to that of a larger site, and that this particular kind of assemblage is characterized both by a patchy distribution of the bone fragments and by a relatively high incidence of trampling.