The present research is focused on developing a novel, scaled-down approach based on the liquid droplet-powder bed interactions for the estimation of an optimal liquid to solid ratio (L/S) required in the twin screw wet granulation. A single droplet ‘powder-bed-nucleation’ process using compacted powder beds of α-lactose monohydrate (soluble) and microcrystalline cellulose (insoluble, swells upon water addition) with varying proportions of hydroxypropyl cellulose, was used to produce wet lumps of powder called ‘powder bed nuclei’. The knowledge from such nuclei formation, droplet weight and its penetration time into the powder bed and weight and hardness of nuclei was used to develop a novel method allowing the estimation of optimal amount of liquid required to produce granules with desired particle size distribution using twin screw granulator. The estimated and the experimental values for L/S showed an acceptable agreement.