Practical optimization of the 4D [(1)H, (13)C, (13)C, (1)H] HMQC-NOESY-HSQC experiment in terms of distribution of resolution over the indirect dimensions is analyzed in detail. Recommendations for an optimal experiment are based on computer simulations assessing the effective resolution of the experiment, defined as the percentage of all possible NOE cross peaks that can be assigned unambiguously on the basis of the spectral data alone. Using actual (13)C-(1)H spectra of an 18-kDa chaperone protein, the analysis shows that experiments with the best effective resolution are also among the most sensitive ones. When combined with an efficient aliasing scheme that reduces indirect spectral space 124-fold, a 4D experiment that yields unambiguous assignments for 41% of all possible NOE cross peaks can be recorded in 28 h. A high-resolution experiment, which can be recorded in 8 days, yields 61% unambiguous assignments and can be analyzed more easily using standard NMR display software. The predictions are verified with experimental 4D spectra from which 1850 NOEs (914 long-range) were extracted for the 18-kDa chaperone protein.