The direct detection of weak 13C15N heteronuclear dipolar couplings in magicangle spinning (MAS) NMR experiments is often difficult. A conventional MAS 13C NMR spectrum has spinning sidebands arising from both dipolar and chemical-shift interactions. Two-dimensional NMR experiments which separate dipolar couplings from the chemical-shift anisotropy simplify considerably their determination. Munowitz and Griffin introduced single-sided ( I ) and double-sided (2) two dimensional magic-angle spinning NMR experiments that have proved invaluable in the study of strong r3C‘H dipolar interactions. Unfortunately, when these techniques are applied to the substantially weaker 13C“N coupling, there are usually not enough dipolar sidebands to characterize effectively the dipolar sideband spectrum, even at low spinning speeds. A variation of the Munowitz and Griffin double-sided experiment, called odd p this yields more dipolar sidebands. The extra sidebands often contain enough information to allow a good determination of weak 13C“N dipolar coupling. Typically, however, the experiment must be done at low spinning speeds of the order of 500 Hz with the consequence that an overwhelming number of rotational sidebands can appear in the chemical-shift dimension. This is undesirable since the dipolar-modulated magnetization is actually measured by summing over sidebands in the chemical-shift dimension. The characterization of weak 13C15N couplings at slow spinning speeds in a complicated carbon spectrum can be simplified by the use of double-cross polarization (4) (DCP) as a selective filter; however, DCP is technically difficult to execute and generally does not give equal selection to spins with different spatial orientations (3). This Communication presents preliminary results on new experiments designed to measure weak heteronuclear dipolar couplings without the problems associated with ODRSE. Weak couplings can be measured at higher spinning speeds; very weak couplings can be measured at lower spinning speeds. Our experiments are extensions of rotational echo double resonance (5) (REDOR) which, in turn, was derived from spin-echo double resonance (6, 7) (SEDOR) . Extension of the dipolar evolution time in rotating solids is the principal feature of the experiments; accordingly, this class of experiments is called extended dipolar modulation (XDM) . XDM is a REDOR experiment and is specified by the number