The variability of quartz hazard stands as one of the most puzzling issues in particle toxicology, notwithstanding the fact that silicosis, the most ancient occupational disease, was the very topic from which the study of the toxicity of particulates developed. Over the years, other adverse effects of silica particles (i.e., lung cancer and autoimmune diseases) were detected and described. However, a few gaps are still present in the physicochemical determinants and cellular pathways involved in the mechanisms of silica pathogenicity. In this perspective, we illustrate how pooling together studies in occupational health and nanotoxicology might fill such gaps, yielding a consistent picture of what imparts toxicity to a given silica source. Recent investigations have shown that crystallinity is not implied in the pathogenic process of silica per se, while patches of disorganized silanols at the surface of both crystalline and amorphous particles can promote membrane damage and inflammation, a process at the origin of silica-related diseases. Introducing these new findings into the accepted multistep model of silica pathogenicity, we obtain a picture of the chemical features of silica governing each cellular step in agreement with the outcomes of major previous studies. We ascribe the origin of the variability of silica hazard mainly to the distribution of various moieties at the particle surface, with silanols playing the major role. Toxicity turns out to be likely predictable by an ad hoc surface characterization. Tailored modifications of the surface can be envisaged to prepare safe materials or blunt toxicity in existing ones.