Heavy metal toxicity is a worldwide problem which is associated with the metal’s high affinity for thiolate rich proteins. Despite the tremendous toxicity concern, the mode of binding of As(III) and Pb(II) to proteins is poorly understood. To clarify the requirements for toxic metal binding to metalloregulatory sensor proteins such as As(III) in ArsR/ArsD and Pb(II) in PbrR or replacing Zn(II) in δ-aminolevulinc acid dehydratase (ALAD), we have employed computational and experimental methods examining these heavy metals binding to designed peptide models. The computational results show that the mode of coordination of As(III) and Pb(II) is greatly influenced by the steric bulk within the second coordination environment of the metal. The proposed basis of this selectivity is the large size of the ion and, most important, the influence of the stereochemically active lone pair in hemi-directed complexes of the metal ion as being crucial. The experimental data show that switching a bulky leucine layer above the metal binding site by a smaller alanine residue enhances the Pb(II) binding affinity by a factor of five supporting experimentally this lone pair steric hindrance hypothesis. These complementary approaches demonstrate the potential importance of a stereochemically active lone pair as a metal recognition mode in proteins and, specifically, how the second coordination sphere environment affects the affinity and selectivity of protein targets by certain toxic ions.