The glauconitic sands in the upper part of the lower‐middle Miocene Berchem Formation are subdivided into the Kiel and Antwerpen Members. Although lithological differences between both members are well known from temporary outcrops in the Antwerp city area, they are difficult to distinguish in boreholes, which hinders regional mapping of these units. In this study, we investigate whether both members can be distinguished on cone penetration tests (CPTs). For this purpose, we correlated multiple outcrops—in which the Kiel and/or Antwerpen Members have been identified—with nearby CPTs. On the CPTs, the boundary between the Kiel and Antwerpen Members is clearly identifiable as it coincides with an abrupt upwards decrease in cone resistance ( q c ). The lower q c of the basal part of the Antwerpen Member is probably related to the finer grain size with more clayey admixture compared to the underlying Kiel Member. This change to a finer grain size is caused by a decrease in depositional energy and sedimentation rates as the region was transgressed during the eustatic sea‐level rise at the start of the Mid‐Miocene Climatic Optimum. On the CPTs, several spikes in q c values were observed within the Antwerpen and Kiel Members. These spikes could be correlated to shell beds, three horizons with sandstones and possibly a hardground. The sandstones appear to be discontinuous, whereas some of the shell beds could be traced across the entire study area. Most shell beds probably represent storm deposits within an otherwise relatively low energetic sedimentary environment. A phosphatic shell bed above the base of the Antwerpen Member is interpreted as the maximum flooding surface, lying in a zone with the lowest q c values for the Antwerpen Member, which might reflect maximum fining. The shell beds and interlayered sands of the Antwerpen Member thin in a southern direction, indicating reduced accommodation space in this direction.