Despite tremendous technical advances in drug discovery, de novo development of small molecule drugs is still challenging. High-throughput screening (HTS) with libraries of natural products and other complex molecules remains the bedrock approach. However, HTS is unsatisfactory in many ways: extraordinary cost, poor efficiency, rampant false positives and a complexity of “hits” that hinders hit-to-lead development. Fragment based drug discovery (FBDD) was brilliantly conceived to overcome these limitations, but has arguably not performed as hoped. The limited impact of FBDD is because most fragment “hit” molecules are very weak binders and are undetectable by current assay methods. The enormous potential of FBDD is therefore lost. Here, an approach is to be developed that can reliably detect weak but specific binding with the goal of helping to reinvigorate and enhance early phase small molecule drug discovery. Faithful detection of binding requires that the ligand and protein concentrations be at least on the order of the dissociation constant, which is practically and financially unrealistic for weak binders. The strategy to remove this basic barrier is simple. The water core of the reverse micelle (RM) is used to confine a single protein molecule and fragments at high enough concentrations to overcome the unfavorable binding entropy. NMR spectroscopy then permits site-resolved detection and quantification of binding affinity at reasonable cost. The first application of RM NMR FBDD highlights its potential to greatly expand small drug discovery. A rule- of-three (Ro3) fragment screen of interleukin-1β (IL-1β) shows that 1) weak yet specific binding can be efficiently detected in a structural context; 2) achieving the required high protein and ligand concentrations is economically feasible; 3) a high hit rate is observed; 4) surface coverage is extraordinary and gives unprecedented connectivity potential; 5) highly desired more polar binders are illuminated. The door is now open to more fully realize the tremendous promise of FBDD but critical questions remain: Is the IL-1β surface coverage typical? What is the distribution of fragment hit affinities of Ro3 and rule-of-five (Ro5) libraries more generally? What are the chemical characteristics of useful fragments to choose for an optimal RM NMR screening library? How useful are the very weakly binding hits for lead development? Does the Ro5 library offer a better compromise of hit affinity and surface coverage? What is the most efficient way to carry out RM NMR screening? Is RM NMR screening quantitatively reliable? This project will address these and other technical challenges that stand in the way of creating a strategy that more fully enables the brilliant insights of the FBDD paradigm and unleashes its originally anticipated potential.