Risk-based approaches are a feature of the modern quality management system. A method of optimization of product quality inspection plan by the risk of non-conformity slippage is proposed. The method is based on a risk ranking matrix, criteria of the failure mode and effects analysis (FMEA), block classification of inspection plans, approaches to non-conformity prediction, and probability multiplication theorem for independent events. The risk of non-conformity slippage was defined as a criterion of inspection plan optimization. The proposed method allows determining the acceptability of the risk, with 100 % quality inspection, in case of abandoning the inspection operation, the possibility of applying sampling and minimum sampling volumes necessary to ensure an acceptable risk level. Relationships were derived to determine the minimum required number of inspected units out of 1,000, with an acceptable risk level in product quality inspection. The initial data for the calculation are the main characteristics of the inspection plan: the probability of the object conformity with the requirements for the controlled quality characteristic, the probability of not detecting non-conformity with the provided inspection method, the rate of non-conformity slippage, which ensures an acceptable risk level. The formula allows calculating the minimum sampling volume that provides an acceptable level of non-conformity slippage risk during the implementation of the product quality inspection plan (QIP). The proposed method was tested on the inspection plan for welds of air tanks of the railway car braking system. It is possible to abandon the original 100 % inspection plan and apply sampling, which provides an acceptable level of non-conformity slippage risk. This allows reducing the volume and costs of inspection by 18 %.