An alternative approach for risk assessment of nongenotoxic substances, the benchmark method, has been evaluated and applied to trichloroethene as a test case. The benchmark dose is the dose that corresponds to a specific increase in risk, normally 1 or 10%. Experimental data from the literature on trichloroethene were used for these calculations. Eighty sets of data on effects on liver, kidney, the central nervous system, and tumors were analyzed. All non-observed-effect levels (NOELs) were higher than the benchmark dose corresponding to 1% extra risk, and 42% of the NOELs and 93% of the lowest-observed-effect levels (LOELs) were higher than the benchmark dose corresponding to 10% extra risk. The present study confirms that the benchmark methodology gives a more detailed picture of dose-response relationships than risk assessment using the NOEL/LOEL approach and facilitates comparison between various toxicity studies. However, the polynomial regression models used in the present study quite often failed to fit the experimental data. Despite the advantages with the benchmark approach, several factors must be considered in the risk assessment process. In the case of trichloroethene, a revised risk assessment using the benchmark approach would lead to a similar guideline value as the traditional NOEL/LOEL approach.