The biograded age estimates calculated from a morphometric index based interpolation procedure using the Rosita (Contusotruncana) lineage have been used to document the geology of deposits of the Pyrenees in the late Cretaceous. These estimates show geological inconsistencies whose size is greater than the precision (relative uncertainty) assumed by the theory. For example, in the sub-Pyrenean basin, a lithological interface was said to be synchronous according to the biograded age estimates: in contrast ammonites put this interface into three different biozones, covering several Ma, depending on the locality. In the Baie de Loya succession (Basque country), the biograded age estimates would indicate a continuous depositional sequence; in fact, the tectonic structure and calcareous nannofossil data prove that the succession has been duplicated, each repetition containing biozones several m.y. long. In the Coniacian-Santonian stages in the eastern Pyrenees, biograded age estimates have been published; comparison with the ammonite fauna whose North American equivalents have been radiometrically dated, shows inconsistencies of 1 to 3 Ma. Finally, the use of the interpolation theory for sequence interpretation of the Aquitaine deposits (sub-Pyrenean basin and the reference-section at Tercis-les-Bains) leads to inconsistencies which are difficult to explain: (a) a sequence boundary has been given different dates (for a single reference calibration using the same fossils at the same place) with no explanation; (b) biograded estimates of age suggest a diachronism of the internal systems-tracts within homologous sequences with synchronous external limits which seems difficult to accept; (c) "changes" in dates larger than the assumed precision of the interpolation procedure as published from one paper to another, but without explanation. The variation in the published dates for the top and bottom of a unit may be greater than the duration of the unit itself. These facts establish that the precision and reproducibility of biograded age estimates have been fundamentally over estimated. The variations suggest that the method is unappropriate for solving problems of age. In fact, we suspect that we have demonstrated that the published biometric indices are influenced by factors other than time alone.