Raman spectroscopy is an optical technique based on the inelastic scattering of monochromatic light that can be used to identify the biomolecular composition of biological cells and tissues. It can be used as both an aid for understanding the etiology of disease and for accurate clinical diagnostics when combined with multivariate statistical algorithms. This method is non-destructive, potentially non-invasive and can be applied in vitro or in vivo directly or via a fiber optic probe. However, there exists a high degree of variability across experimental protocols, some of which result in large background signals that can often overpower the weak Raman signals being emitted. These protocols need to be standardised before the technique can provide reliable and reproducible experimental results in an everyday clinical environment. The objective of this study is to investigate the impact of different experimental parameters involved in the analysis of biological specimen. We investigate the Raman signals generated from healthy human cheek cells using different source laser wavelengths; 473 nm, 532 nm, 660 nm, 785 nm and 830 nm, and different sample substrates; Raman-grade calcium fluoride, IR polished calcium fluoride, magnesium fluoride, aluminium (100 nm and 1500 nm thin films on glass), glass, fused silica, potassium bromide, sodium chloride and zinc selenide, whilst maintaining all other experimental parameters constant throughout the study insofar as possible.