Sediment toxicity tests are used to measure the effects of contaminants in sediment and are important management tools for agencies that regulate dredging and dredged material disposal, wastewater discharges, and other activities that might affect the aquatic environment. The interpretation of the results of routine toxicity tests can, however, be confounded by naturally occurring factors, such as ammonia, sulfides, sediment grain size, and salinity. Ammonia has been suspected for many years to be a sediment toxicant. It is only recently that the sensitivity of routine test organisms to ammonia has been quantified, and even more recently that dose-responses have been related to ammonia concentrations that occur in standard sediment toxicity tests. The first objective of this study was to compare the effects of ammonia in sediment toxicity tests under two pore water ammonia reduction procedures. The first procedure reduced total pore water ammonia from test sediments by replacing overlying water two times in a 24-h period. The second procedure used natural bacterial processes to reduce pore water ammonia. The second objective was to evaluate the current no-effect concentrations (originally based on 96-h water-only toxicity tests) using the overlying- water-exchange procedure. The final objective was to determine whether ammonia concentrations would increase to toxic levels in sediments that had naturally reached no-effect concentrations, but that were then physically disturbed. The marine amphipod, Rhepoxynius abronius, was exposed to sediments with low (≈30 mg/L), moderate (≈ 100 mg/L), and high (≈200 mg/L) initial concentrations of pore water total ammonia in five separate 10-day toxicity tests. One test followed standard protocols (ASTM E 1367-90). Two tests were initiated after pore water ammonia concentrations were reduced by overlying water exchange to ≤30 mg/L and ≤20 mg/L. Two tests were initiated after pore water total ammonia decreased by natural reduction to <10 mg/L, half of the test chambers from the final test were disturbed by shaking before test organisms were placed in the test containers. Toxicity of sediments with higher initial concentrations of pore water ammonia decreased significantly when ammonia was reduced to ≤30 mg/L. Reaching 30 mg/L required up to 12 days with the reduction procedure (24 exchanges of overlying water) or up to 47 days under the natural-reduction procedure. There was no difference in toxicity between the naturally reduced and the disturbed samples.