Noninvasive cell analyses are increasingly important in the medical field. A coherent anti-Stokes Raman scattering (CARS) microscope is the noninvasive imaging equipment and enables to obtain images indicating molecular distribution. However, due to low-signal intensity, it is still challenging to obtain images of the fingerprint region, in which many spectrum peaks correspond to compositions of a cell. Here, to identify cell differentiation by using multiplex CARS, we investigated hyperspectral imaging of the fingerprint region of living cells. To perform multiplex CARS, we used a prototype of a compact light source generating both pump light and broadband Stokes light. Assuming application to regenerative medicine, we chose a cartilage cell, whose differentiation is difficult to be identified by change of the cell morphology. Because one of the major components of cartilage is collagen, we focused on distribution of proline, which accounts for approximately 20% of collagen. The spectrum quality was improved by optical adjustments of the power branching ratio and divergence of Stokes light. Periphery of a cartilage cell was highlighted in a CARS image of proline, and this result suggests correspondence with collagen generated as an extracellular matrix. The possibility of noninvasive analyses by using CARS hyperspectral imaging was indicated.