Non-uniform weighted sampling (NUWS) is a sampling strategy, related to nonuniform sampling (NUS) in the limit of long acquisition times, in which each indirect increment of a multidimensional spectrum is sampled multiple times according to some weighting function. As the spectrum is fully sampled it can be processed in a conventional manner by the discrete Fourier transform, making the analysis of sensitivity much more straightforward than for NUS data. Previously, 2-3 fold increases in signal-to-noise ratio (SNR) have been reported using NUWS. However, as the sampling schedule acts as a window function, the observed SNR must be compared with uniformly sampled data apodized using the same weighting function. On doing this, we calculate more modest improvements of 10-20% in SNR, and these are verified experimentally for spectra of α-synuclein and YFP. Nevertheless, we prove that NUWS always improves the sensitivity compared with identically processed uniformly sampled data, and when combined with rapid recycling experiments such as the SOFAST-HMQC, NUWS methods have the potential to make a useful and practical contribution to sensitivity-limited measurements.