Trend analysis and forecasting of time series data on air-pollutants is important to design effective measures to minimize damages to ecosystems and human health. In this study, autoregressive, moving average, autoregressive-moving average and autoregressive integrated moving average processes of different order were implemented to examine patterns of depositions and emissions. Analysis was undertaken to examine stationarity of the series or to design a method to create stationary series. The model that satisfied selected statistical criteria was chosen to make forecasts. Forecasts of depositions were compared with critical loads by watersheds. The findings of this study indicated that both wet depositions and emissions of SO2 and NOx data exhibited non-stationarity. After removing non-stationarity, suitable time-series model was selected for short-run forecasting (1994 to 2005). The resulting depositions and emissions data were examined with respect to their long-run movement and critical deposition loadings. The analysis showed that excess wet depositions of SO2 and NO3 would be major problems at least for ten years. Most of these problems are observed in Atlantic Canada and few watersheds in Quebec and Ontario. Although emissions of SO2 have declined, emissions of NOx remained unchanged or increased compared to the 1980 level. Considering the fact that these pollutants contribute to acidification, eutrophication and formation of secondary particulates that are hazardous to human health, it is necessary to find ways of further reducing emissions and depositions of these pollutants. While substantial progress has been made with respect to reduction of SO2 emissions (especially in Canada), the analysis presented in this study indicated that there must be substantially more reductions to ensure the protection of sensitive ecosystems. Thus, evidences similar to those presented in this study should be gathered to initiate negotiations for reductions beyond the 2005 or 2010 commitments.