Dynamic nuclear polarization (DNP) permits a similar to 10(2)-10(3) enhancement of the nuclear spin polarization and therefore increases sensitivity in nuclear magnetic resonance (NMR) experiments. Here, we demonstrate the efficient transfer of DNP-enhanced H-1 polarization from an aqueous, radical-containing solvent matrix into peptide crystals via H-1-H-1 spin diffusion across the matrix-crystal interface. The samples consist of nanocrystals of the amyloid-forming peptide GNNQQNY(7-13), derived from the yeast prion protein Sup35p, dispersed in a glycerol-water matrix containing a biradical polarizing agent, TOTAPOL. These crystals have an average width of 100-200 nm, and their known crystal structure suggests that the size of the biradical precludes its penetration into the crystal lattice; therefore, intimate contact of the molecules in the nanocrystal core with the polarizing agent is unlikely. This is supported by the observed differences between the time-dependent growth of the enhanced polarization in the solvent versus the nanocrystals. Nevertheless, DNP-enhanced magic-angle spinning (MAS) spectra recorded at 5 T and 90 K exhibit an average signal enhancement epsilon approximate to 120. This is slightly lower than the DNP enhancement of the solvent mixture surrounding the crystals (epsilon approximate to 160), and we show that it is consistent with spin diffusion across the solvent-matrix interface. In particular, we correlate the expected DNP enhancement to several properties of the sample, such as crystal size, the nuclear T-1, and the average H-1-H-1 spin diffusion constant. The enhanced H-1 polarization was subsequently transferred to C-13 and N-15 via cross-polarization, and allowed rapid acquisition of two-dimensional C-13-C-13 correlation data.