The article presents a critical review of theoretical and methodological concepts upon which analytical works about Amazonian peasantries since the 1960s have been based. The vision of Amazon peasants as shifting cultivators with low productivity who destroy their ecological habitat and are condemned to disappear due to the advance of large landowners (the model of the frontier cycle) is contrasted with the tendency for the consolidation of family agriculture based on more complex production systems, including permanent cultures, small animal husbandry and cattle. This tendency is more clearly evident in the Northeast of Pará, but has also been statistically confirmed for the State of Pará and the Northern Region of Brazil. This means that the hypothesis of the frontier cycle has limited validity for older colonization regions. However, recent research on borders showed production systems which concentrate on cattleraising in the South of Pará and on permanent cultures in the Transamazônica region. These systems cannot simply be classified as shifting cultivation but represent different trajectories from that of the Northeast of Pará.
