This study tests the short-term forecasting improvement afforded by the inclusion of low-frequency inputs to ANN R-R models that are first optimized by using only fast response components: stream flow and rainfall. Ten low-frequency input candidates are confronted: the potential evapotranspiration, the antecedent precipitation index (APIi, i=7, 15, 30, 60, and 120 days) and a proposed soil moisture index (SMIA, A=100, 200, 400 and 800 mm). The APIi are non-decayed moving average precipitation series, while the SMIA are calculated through the soil accounting reservoir of the lumped conceptual rainfall-runoff model GR4J. Results, based on the Serein and Leaf Rivers, reveal that only the SMIA time series are useful to the one-day-ahead stream flow mapping. Both the potential evapotranspiration and the APIi time series fail to improve the ANN performance.