Critical Load (CL) methodology is currently used throughout Europe to assess the risks of ecological damage due to sulphur and nitrogen emissions. Critical acid neutralising capacity (ANC CRIT ) is used in CL estimates for freshwater systems as a surrogate for biological damage. Although UK CL maps presently use an ANC value of 0 meq l 1 , this value has been based largely on Norwegian lake studies, in which brown trout is chosen as a representative indicator organism. In this study, an ANC value specific for brown trout in Scottish streams was determined and issues were addressed such as salmon and trout sensitivity in streams, episodicity, afforestation and complicating factors such as dissolved organic carbon (DOC) and labile aluminium (Al-L). Catchments with significant forest cover were selected to provide fishless sites and to provide catchment comparisons in unpolluted areas. Chemical factors were the primary determinant with land use a secondary determinant of the distribution of salmonid populations at the twenty-six study sites. ANC explained more variance in brown trout density than pH. The most significant index of episodicity was percent of time spent below an ANC of 0 meql -1 . An ANC CRIT value of 39 meql -1 was obtained based on a 50% probability of brown trout occurrence. The use of this revised ANC CRIT value in the CL equation improved the relationship between trout status and exceedance of CLs. Uncertainties associated with variations in Al-L at any fixed ANC CRIT , particularly within forested catchments, and the role of DOC in modifying the toxicity of Al-L are discussed.