We used multiple linear regression analyses to explore empirical relationships between dissolved organic carbon (DOC) concentrations, weather, and acidification in long-term data sets from 12 small Boreal Shield lakes in Ontario, Canada. In two lakes in which pH changes have been very large (4.5 to ∼6.0), pH explained most of the temporal variation in DOC concentrations. In the remaining lakes, long-term average previous temperature (on the scale of a decade or more) was usually the best explanatory variable for DOC concentrations. Lake-specific multiple regression models constructed from long-term and short-term attributes of weather (long-term average previous temperature and precipitation, winter-spring precipitation, summer precipitation, summer sunshine) and pH explained between 41% and 96% of the temporal variation in DOC concentrations during the entire monitoring period for these lakes (n = 16 26 years). Multiple regression models considering only the period common to all lakes, 1987 to 2003 (n = 16-17 years), explained 35%-96% of the variation in DOC concentrations. The importance of long-term and short-term attributes of weather in explaining temporal variations in DOC concentrations suggests that changes in climate will have large effects on lake clarity; however, the interactions between weather-related effects may be very complex. © 2008 NRC.