Soil samples were collected at 15-cm increments to a depth of 75 cm from plots on a silt loam soil where until several years earlier and for 14 years, anaerobically digested sewage sludge had been annually applied by furrow irrigation. The study protocol consisted of four replications of 6.1 x 12.2-m plots with 0 (T0), 1/4-maximum (T1), 1/2-maximum (T2) and maximum (T3) sludge application rates randomized within blocks. When sludge applications were terminated, maximum sludge-treated plots had received 765 Mg ha-1 (dry weight equivalent) of sludge solids. Total soil concentrations of Cd, Cr, Cu, Ni, Pb and Zn had been significantly enhanced by all sludge application rates to a soil depth of 30 cm. Below the 30-cm depth, total soil Cd was increased to 75 cm, total Zn to 45 cm (T2 and T3 only), total Cr to 60 cm (T2 and T3 only), but total Cu, Pb, and Ni were not increased at depth. Despite the lack of significant increases in subsoil concentrations for some metals, mass balance calculations showed a relatively high proportion of all the above sludge-borne heavy metals to be unaccounted for in the soil profile for each application rate. Mass balance calculations of losses ranged from a high of 60% for Ni to a low of 36% for Cu and Pb. Similar losses were calculated from metal concentrations measured in soil samples taken at the time the sludge was applied. In soil surface samples (0-15 cm) from maximum sludge-treated plots, percentages of total metal concentration extracted with 4.0 M HNO3 ranged from a low of 31 for Zn to a high of 75 for Cu. Efficiency of metal extraction by HNO3 was inconsistent, depending on the soil horizon and sludge treatment, so that evaluation of HNO3-extractable metals is not a reliable method of estimating total metal retention in the profiles. In soil surface samples from maximum sludge-treated plots, the percentage of total metal contents extracted with DTPA ranged from a low of 0.03 for Cr to a high of 59 for Cd. The DTPA extractable levels of Cu, Ni, and Pb were higher in the subsoils of the sludge-treated soils, indicating that these metals had been redistributed from the surface layer to deeper zones in the profile of sludge-amended soil, despite the absence of elevated total concentrations of these three metals in the deeper subsoil.