Synthetic model soils were used to evaluate redistribution of Pb in both the three-stage sequential extraction procedure proposed by the Commission of the European Communities Bureau of Reference (BCR, now Standards, Measurements and Testing Programme) (1993) and the commonly-applied procedure of Tessier et al. (1979). Redistribution of Pb in both schemes was found to be substantial, sufficient to render meaningless any attempts to quantify the original phase associations of Pb in cases where these are not known. In the model systems, the manganese oxide (hausmannite) and humic acid phases were found to be responsible for most of this redistribution in both schemes. Despite the fact that both schemes allow severe Pb redistribution to occur, differences are apparent in the manner in which this occurs. Whereas Tessier et al.’s procedure in all cases yielded the highest recovery of Pb at the reducing step, the three-stage BCR procedure released the most Pb at the oxidizing step, in both cases irrespective of where it originated. These differences are attributable to a combination of high temperature (96 °C) and the presence of 25% v/v acetic acid in Tessier et al.’s reducing step. These conditions were found to reduce the sorptive capacity of humic acid by more than 50%, in addition to retaining more of the Pb in solution by complexation with acetate ions. Extensive redistribution in the three-stage BCR procedure was also observed in systems where natural soils were mixed with Pb-spiked synthetic phases. In keeping with the results relating to synthetic soil models, the organic matter and Fe/ Mn oxide fractions of the natural soils appear to dominate the redistribution of Pb during the extraction. Two approaches were tested in terms of their ability to counteract the redistribution of Pb. Each approach was applied to the first extraction step of the BCR scheme, which underwent a range of modifications. These involved the use of 6 distinct solid sequestering materials, and 5 dissolved complexing agents. The most promising of all the approaches are those which involve the two soluble structurally-distinct complexing agents cryptand 2.2.2 (a macrobicyclic ligand) and nitrilotriacetic acid (NT A). The addition of each ligand generally improved the percentages of Pb recovered from the multi-phase model soils, with Pb originally spiked on calcite. However, the effectiveness of the cryptand and NTA at inhibiting redistribution (at concentrations of 9 x 10⁻³ and 1 x 10⁻⁴ M, respectively) was dependent on the percentages (by weight) of the hausmannite and humic acid in each model system. In the presence of no more than 2% hausmannite, both cryptand and NTA could satisfactorily recover about 60% of the Pb, provided that the percentages of humic acid in the systems were not higher than 10% and 5%, respectively. The preliminary criteria for selecting a potentially useful soluble complexing agent suggested by the results of this study are that the ligand possess (a) a high stability constant for the ion involved, and (b) a semi-encapsulated metal binding site (which appears to inhibit direct surface extraction from occurring). It is desirable in the case of Pb to decrease the extraction time of the first step in the BCR scheme to minimize opportunity for redistribution. The modified extraction solution using the cryptand successfully recovered Pb from a mixture of Pb-spiked calcite and a “clean” natural soil. However, when applied to contaminated roadside soil and dust samples (in the absence of a spiked phase), the modified extraction solution did not significantly alter the recovery pattern of Pb, relative to the conventional extraction solution. This result may reflect the natural distribution of Pb in these samples. Four parks in the City of Hamilton were selected as monitoring sites for assessment of levels of Pb in roadside soil and street dust during the removal of leaded-petrol from the New Zealand market. Samples were collected every 2 - 3 months during November 1995 to September 1996. The mean concentrations of Pb in top 2 cm kerbside soil from all the parks reveal a slight downward trend (statistically significant in three parks) whereas those of the fine and coarse fractions of kerbside dust were oscillating.