The detection of molecules that can bind to active sites of protein targets and the measurement of their affinities is a promising application of NMR. Nowadays, the screening of drug candidates is routinely done by NMR in pharmaceutical industry. We have proposed to use the relaxation of Long-Lived States (LLS) for drug screening by NMR. Long-lived states are nuclear spin states whose decay time constant Tlls can be much longer than the longitudinal relaxation time T1. LLS can be used to screen and determine the dissociation constant KD of molecular fragments that bind weakly to protein targets. The use of LLS for fragment screening leads to a spectacular increase in contrast between free and bound ligands, and thus allows one to characterize binding of fragments with very weak affinities, with KD in the millimolar range, which is difficult to achieve by other methods such as ITC. By exploiting the LLS behavior of a spy molecule, we experimentally demonstrate that it is possible to measure dissociation constants KD as large as 12 mM, corresponding to very weak binding, where most other biophysical techniques fail, including other NMR methods based on the observation of ligands. Furthermore, we have combined LLS for screening for improved contrast with 1H dissolution-DNP to enhance the sensitivity. DNP-enhanced screening for measuring LLS signals of a weak ligand allows one to use very low concentrations of ligands and proteins. We observed dramatic differences between the spectra of the ligand in the presence or absence of a protein, or in the presence of the protein combined with a stronger ligand. Moreover, we have explored LLS involving pairs of 19F nuclei to study binding phenomena. Indeed, fluorine detection is quite interesting because it offers the possibility to perform screening experiments without any problems due to overlapping signals. In a customdesigned fluorinated ligand that binds trypsin, we have observed a promising ratio Tlls / T1 > 4. This fluorinated ligand has been used as spy molecule in competition experiments, which allowed us to rank the affinities and estimate dissociation constants of arbitrary ligands that do not contain any fluorine.