In situ air voids are a function of mix design (aggregate type and gradation, and bitumen content), manufacture and the level of compaction achieved during construction and subsequent traffic [1-2]. During the improvement of N12-19 highway construction project between Snake road interchange and Kingsway interchange in Johannesburg, South Africa, the quality control (QC) of a bitumen treated base layer was done according to Committee of Land Transport Officials (COLTO) specifications where compliance is a statistical judgment of three parameters namely, the relative compaction, binder content (lab binder) and voids in the mix (lab air voids).

The focus of this paper is to analyse only one of the three parameters namely, voids content obtained from the laboratory and from in-situ field measurements. It was found that for a given random sample, there is a considerable difference between in-situ void and lab void for the same material. Therefore, it becomes interesting to statistically evaluate the random sample results of in-situ air voids in order to decide on its compliance with quality control requirements. The study concluded that QC evaluation based on lab void results might be accepted but use of in-situ air voids values may lead to rejection decision, for the same bituminous mixtures.