Short-term forecasting is increasingly used to support operational decision-making in hydropower and water resources systems. While many forecasting approaches have been developed for large reservoirs, their application to run-ofriver and mini hydropower schemes remains challenging due to differences in scale, hydrological response, and data availability. These systems typically have limited storage, rapid response to rainfall, and shorter decision-making windows, which constrain forecasting effectiveness. This paper discusses key challenges encountered during the development of a forecasting model for a run-of-river mini hydropower system, with reference to the Kota 2 Mini Hydro Project. The project is located in the northern region of Sarawak, approximately 40 minutes from Lawas town, within a generally high-elevation catchment, and comprises a run-of-river scheme with a weir crest level of 442.5 m above sea level and an installed capacity of 10.5 MW, operating in a flashy river system. The study focuses on practical considerations that influence model development and operational use, rather than emphasising predictive performance. The discussion considers how data characteristics, the use of rainfall forecast inputs at small catchment scales, temporal alignment between forecast inputs and system response, and hydrological dynamics typical of run-of-river systems influence forecasting effectiveness. Based on the lessons learned, the paper outlines practical considerations for improving forecasting model development for run-of-river schemes. These include prioritising data reliability, balancing model complexity with robustness, and aligning forecast outputs with operational decision-making needs. The insights presented aim to support more effective and realistic forecasting applications for run-of-river mini hydropower systems.