Variation in mill feed size distribution impacts on Autogenous/Semi-autogenous (AG/SAG) grinding mill operation, influencing mill disturbances and performance. Under batch dry and wet grinding conditions using a purpose-built laboratory-based AG/SAG mill and acoustic sensor, the present work investigates the feasibility of a real-time monitoring technique of different mono-sized ore feed fractions and binary mix ratios of distinct feed size fractions., Simple feed size variations estimation models were developed by signal root mean square analysis derived from the time-domain acoustic signal. The results obtained delineate that mill noise emissions generally increases with an increase in ore feed size. The mix ratios of relatively coarse and fine feed sizes indicated that mill acoustic response was dampened with an increase in the fine size fractions. The more suitable fit correlation plot of both mono-sized and hetero-sized feed fractions (both dry and wet grinding environment) and acoustic emissions showed a nonlinear relationship. Wet milling in the presence of steel balls and water showed a more linear correlation compared with only rock milling. The results suggest that mill acoustic emissions are sensitive to mill feed variation with a promising prospect towards real-time monitoring and optimisation of AG/SAG mill feed size.