Over the period 1973 to 1992 spring flows in the Skoga River, Southern Iceland experienced a statistically significant decline. Moreover, over the same period there was a switch between periods of above, below and above average discharge for the years 1973-1979, 1980-1990 and 1991-1992 respectively. This study attempts to uncover the regional climate and North Atlantic atmospheric circulation controls on the secular trend of spring discharge and the low frequency variability of annual discharge. Regional climate is described by monthly temperature and precipitation from Vik, Southern Iceland while atmospheric circulation is characterized by the North Atlantic and Arctic Oscillation indexes and meridional and zonal vapour flux components over Iceland. It is shown that spring discharge decline is attributable to spring cooling which is linked with a strengthening of the April NAOI and a weakening (but statistically insignificant) of the southwesterly circulation over the study area. The observed cooling of spring months has also been observed for West Greenland and is opposite in sign to that for the Northern Hemisphere. Low frequency variations in discharge match those of the large-scale atmospheric circulation, especially the direction and magnitude of its meridional and zonal components over Iceland. Of great interest is the rapid rise of discharge levels towards the end of the study period. These coincide with a rapid negative to positive phase shift of the NAO and AO. This phase shift also corresponds to the beginning of the period of rapid warming in observed surface air temperatures over Arctic latitudes. Both the NAO and AO also show possible quasi-periodic behaviour in their relationship with discharge. Moreover, for the majority of the study period, NAO discharge associations are weakly positive, but the AO displays a tendency towards a stronger inverse forcing of discharge. Furthermore, coincident with the rapid transition of both the NAO and AO to a strong positive phase, the strength of NAO/AO discharge relationships has abated. Findings are discussed in relation to water resource forecasting. A conceptual model describing Southern Iceland discharge regional climate and atmospheric circulation links is also presented as a basis for the development of a hydroclimate prediction system.