The aggregation process of alpha-hANP has been investigated in vitro at physiological concentrations by gel chromatographic procedures using a radiolabeled tracer incubated in PBS and in plasma. In PBS big forms of ANP are organized as a peak eluting from both Sephacryl S-100 and S-300 HR in the void volume of the columns; in plasma, besides this major peak, a second radioactive peak is evident, eluting from Sephacryl S-100 HR around the HSA region. After gel chromatography on Sephacryl S-300 HR the major peak appears to consist of three components of different molecular size. Some information about the nature of these peak materials comes from the result of parallel incubations of partially aggregated (seed or nucleus) and aggregate depleted tracer. The comparison between the two time courses of big ANP formation indicates that: (a) ANP aggregation is a nucleation-dependent process, with a lag time longer than 8 days, at picogram peptide levels and (b) the aggregated forms of peptide are those eluting in the void volume, the other plasma peaks being probably expression of a binding, neither saturable or reversible, to some plasma components. The principle of seeded polymerization, used to detect ANP aggregates present in the plasma, indicates that: (a) the endogenous big ANP cannot act as a nucleus for polymerization and it likely consists of non-fibrillar ANP aggregates and/or bound ANP, and (b) this experimental approach can be suitable to evidence ANP binding plasma factors for further characterization studies.