In the present paper the authors studied theoretically and experimentally the relationship between spectral resolution and grating density, the limitations to improve the spectral resolution by using high density grating, the use of longer focal length grating to increase spectral resolution without compromising instrument throughput and the effect of slit width on spectral resolution and sensitivity. Finally, two experiment results were provided to show why higher spectral resolution is important to ensure that critical information is not lost during a Raman measurement. Stressed silicon was produced by growing a thin crystalline layer of Si on an Si(x)Ge(1-x) substrate. It is possible to use Raman spectroscopy to probe the stress in the Si(x)Ge(1-x) and Si layers at the same time. The parameter to monitor the stress is the position of the Si-Si vibrational mode in Si(x)Ge(1-x) and Si. Such a measurement requires high spectral resolution because the peaks exhibit very subtle shifts. The authors' results clearly demonstrate that the resolution offered by a high density grating is necessary to properly monitor the very small frequency shift of this cap-layer Si-Si mode in order to properly characterize the strain structure. The Raman band around 180 cm(-1) is assigned to the radial breath mode of single wall carbon nanotube (SWCN). By measuring the frequencies excited with different laser, the diameters of the sample can be obtained. Practically, sample is always composed of SWCN with different but very close diameters and their Raman bands might overlap together and are difficult to determine the frequencies. The authors' results showed that only higher resolution with the long focal length spectrometer can give accurate number and frequencies of Raman bands, which leads to a correct analysis of the diameter distribution.