A series of bioassays was carried out to assess the acute toxicity of neutralized, filtered, bleached kraft pulp mill effluent (BKME) from a single mill to underyearling Pacific salmon in fresh soft water, at 10–13 C. Toxicity expressed in terms of the 96-hr LC50 varied from 22% of full strength BKME to nontoxic in different collections. A procedure is described for estimating the 4-day LC50 from geometric mean survival time data. Toxicity of effluents changed unpredictably with storage (even at 2 C), and declined with air stripping. Comparison of continuous flow and static test procedures indicated that continuous flow procedures yield somewhat higher toxicity results than static tests. Experiments with varying fish densities indicate that measurable toxicity is less in static tests with heavy fish loading. Use of loading densities of 2.5 liters/g fish or better is recommended. Young sockeye salmon (Oncorhynchus nerka) appeared most sensitive, and pink (O. gorbuscha) and coho (O. kisutch) salmon somewhat more resistant to toxic BKME solutions. No correlation was found between time to death and condition factor in the size range of underyearling coho tested (3.0–7.3 cm). Recommendations are made for routine and regulatory bioassay procedures.