A 6 in. (15 cm) bentonite in the Z coal (Cretaceous–Tertiary boundary) in eastern Montana was sampled at four different places, and biotite, sanidine, and zircon were separated from the clay. U–Pb analyses of purified zircons yielded small systematic variations from concordant U–Pb dates. Plotting the data on a concordia diagram, a short linear discordia line intersects the concordia at with an MSWD of 1.16. The systematic variation of the four sets of zircon U–Pb data on the concordia plot may be an artifact of the sampling and purification procedure, or could result from natural sample variation from minor contamination. Biotite fractions of varying specific gravity were obtained for each of the four Z coal bentonite samples and (together with the matching purified sanidine fraction) were analysed for Rb–Sr dating. Excluding those lighter biotite fractions found to have lost 30% or more of their original Rb, an isochron was obtained giving an age of 63.7 ± 0.3 Ma with an initial 87Sr/86Sr ratio of 0.7061 ± 1 and an MSWD of 1.07.To investigate further the Rb–Sr variations in altered bentonite biotite, a large biotite sample from the Ordovician Kinnekulle A1 bentonite of southwestern Sweden was separated into 11 fractions of decreasing specific gravity. Rb–Sr analysis of these fractions also showed a departure from a linear isochron when more than about 30% of the original Rb had been lost. Chemical analysis and X-ray diffraction revealed the biotite-alteration process to be vermiculitization, but gave no definite reason why biotites that retain more than 70% of their original Rb give usable Rb–Sr data. Though some of the alteration may have taken place when the bentonite was deposited as an ash, most of the alteration probably occurred in recent times. The Kinnekulle A1 bentonite Rb–Sr isochron for biotite and sanidine gives an age of 447 ± 1 Ma with an initial 87Sr/86Sr ratio of 0.7094 ± 0.0003 and an MSWD of 3.7.