The presence of naturally occurring asbestos (NOA) in many areas worldwide requires an enhanced geological risk evaluation to ensure workplace safety from asbestos during large construction projects. Due to the complexity of the geological risk definition, health and safety regulations for working with asbestos-bearing materials are often not enforceable in NOA settings. Therefore, to correctly estimate the risk of NOA in these scenarios, new procedures are urgently needed to provide (1) a detailed geological model representative of the possible presence of the asbestos, (2) representative sampling, and (3) a reliable quantitative determination of asbestos content in rocks. This work aims to discuss the improvements on the two latter points specifically developed during the design of the “Gronda di Genova” project, a 50-km-long tunnel bypass partially designed in the NOA-bearing meta-ophiolites of the Ligurian Alps and ophiolites of the northern Apennines in Italy. Implementation of Gy's theory on sampling was used to maintain statistical validity during sample processing from the primary sample to the analytical sample and is here described. The scanning electron microscopy/energy dispersive spectroscopy procedure for the quantification of NOA was improved with an error analysis delivering the minimum number of fibers to be measured to achieve the best analytical results.