A carbon-detected TROSY-optimized experiment correlating H-1(N), N-15, and C-13' resonances, referred to as c-TROSY-HNCO is presented, in which the H-1(N) and N-15 TROSY effects are maintained in both indirect dimensions, while the directly detected C-13' is doubly TROSY-optimized with respect to H-1(N) and N-15. A new strategy for sensitivity enhancement, the so-called double echo-antiecho (dEA), is described and implemented in the c-TROSY-HNCO experiment. dEA offers sensitivity enhancement of root 2 in both indirect dimensions and is generally applicable to many multidimensional experiments. A carbon-detected HNCO experiment, c-HNCO, without TROSY optimization and sensitivity enhancement is also designed for comparison purposes. Relaxation simulations show that for a protein with a rotational correlation time of 10 ns or larger, the c-TROSY-HNCO experiment displays comparable or higher signal-to-noise (S/N) ratios than the c-HNCO experiment, although the former selects only 1/4 of the initial magnetization relative to the later. The high resolution afforded in the directly detected carbon dimension allows direct measurement of the doublet splitting to extract (1)J(C alpha C') scalar and D-1(C alpha C') residual dipolar couplings. Simulations indicate that the c-TROSY-HNCO experiment offers higher precision (lower uncertainty) compared to the c-HNCO experiment for larger proteins. The experiments are applied to N-15/C-13/H-2/[Leu,Val]-methyl-protonated IIBMannose, a protein of molecular mass 18.6 kDa with a correlation time of similar to 10 ns at 30 degrees C. The experimental pairwise root-mean-square deviation for the measured (1)J(C alpha C') couplings obtained from duplicate experiments is 0.77 Hz. By directly measuring the doublet splitting, the experiments described here are expected to be much more tolerant to nonuniform values of (1)J(C alpha C') (or (1)J(C alpha C') + D-1(C alpha C') for aligned samples) and pulse imperfections due to the smaller number of applied pulses in the "out-and-stay" coherence transfer in the c-HNCO-TROSY experiment relative to conventional H-1-detected "out-and-back" quantitative J correlation experiments.