Crop models are the primary means by which agricultural scientists assess climate change impacts on crop production. Site-based and high-quality weather and climate data is essential for agronomically and physiologically sound crop simulations under historical and future climate scenarios. Here, we describe a bias-corrected dataset of daily agro-meteorological data for 109 reference weather stations distributed across key production areas of maize, millet, sorghum, and wheat in ten sub-Saharan African countries. The dataset leverages extensive ground observations from the Global Yield Gap Atlas (GYGA), an existing climate change projections dataset from the Inter-Sectoral Model Intercomparison Project (ISIMIP), and a calibrated crop simulation model (the WOrld FOod Studies -WOFOST). The weather data were bias-corrected using the delta method, which is widely used in climate change impact studies. The bias-corrected dataset encompasses daily values of maximum and minimum temperature, precipitation rate, and global radiation obtained from five models participating in the Sixth Phase of the Coupled Model Intercomparison Project (CMIP6), as well as simulated daily growth variables for the four crops. The data covers three periods: historical (1995-2014), 2030 (2020-2039), and 2050 (2040-2059). The simulation of daily growth dynamics for maize, millet, sorghum, and wheat growth was performed using the daily weather data and the WOFOST crop model, under potential and water-limited potential conditions. The crop simulation outputs were evaluated using national agronomic expertise. The presented datasets, including the weather dataset and daily simulated crop growth outputs, hold substantial potential for further use in the investigation of future climate change impacts in sub-Saharan Africa. The daily weather data can be used as an input into other modelling frameworks for crops or other sectors (e.g., hydrology). The weather and crop growth data can provide key insights about agro-meteorological conditions and water-limited crop output to inform adaptation priorities and benchmark (gridded) crop simulations. Finally, the weather and simulated growth data can also be used for training machine learning techniques for extrapolation purposes.