Cell dry mass is principally determined by the sum of biosynthesis and degradation. Measurable change in dry mass occurs on a time scale of hours. By contrast, cell volume can change in minutes by altering the osmotic conditions. How changes in dry mass and volume are coupled is a fundamental question in cell size control. If cell volume were proportional to cell dry mass during growth, the cell would always maintain the same cellular mass density, defined as cell dry mass dividing by cell volume. The accuracy and stability against perturbation of this proportionality has never been stringently tested. Normalized Raman Imaging (NoRI), can measure both protein and lipid dry mass density directly. Using this new technique, we have been able to investigate the stability of mass density in response to pharmaceutical and physiological perturbations in three cultured mammalian cell lines. We find a remarkably narrow mass density distribution within cells, that is, significantly tighter than the variability of mass or volume distribution. The measured mass density is independent of the cell cycle. We find that mass density can be modulated directly by extracellular osmolytes or by disruptions of the cytoskeleton. Yet, mass density is surprisingly resistant to pharmacological perturbations of protein synthesis or protein degradation, suggesting there must be some form of feedback control to maintain the homeostasis of mass density when mass is altered. By contrast, physiological perturbations such as starvation or senescence induce significant shifts in mass density. We have begun to shed light on how and why cell mass density remains fixed against some perturbations and yet is sensitive during transitions in physiological state.