The Classical Least Square (CLS) fitting method was used to analyze the Raman spectra of living cells with the aim of identification of new phenotype-specific spectral markers for osteoblasts. The following chemicals were used for the CLS model: DNA, RNA, serum albumin, chymotrypsin and phosphatidyl choline. In this study we analyzed primary mature osteoblasts as well as two other cell types used as potential sources of osteoblasts: embryonic stem cells and fetal bone cells. The results obtained suggest that the Raman spectra of the cell types can be well approximated with a linear combination of the Raman spectra of the biopolymers used in the CLS model. The relative concentrations of the CLS components varied significantly between cell types, indicating that this analytical method could be used for phenotypic identification of osteoblasts.