Ion probe data for zircons from tuffs within the Llfynant flags (Arenig) and the Serw Formation (lower Llanvirn) of north Wales have been revised using better statistical methods for separating detrital ages, making allowance for recently-found variability in radiogenic 206Pb/238U in the reference zircon SL13, and testing the sensitivity of the ages to the secondary ion discrimination slope. The revised ages are options of 469.2 ± 2.1 (σ) or 472.9 ± 2.9 Ma for the Llfynant flags dependent on mixture modelling, and 465.3 ± 1.4 Ma for the Serw Formation. All ages are within error of previous SHRIMP results and the Serw age now has the same numerical value as a previous MSID age for the same sample. It is shown that an MSID age of 483 ± 0.5 Ma with interpreted Pb loss for a late Tremadoc bentonite is dependent on the correction for common Pb, and that a slightly more radiogenic choice for the common Pb composition places nearly all data on Concordia. The latter would indicate that the bentonite might contain two zircon populations: inherited grains at 482 Ma and tuff magmatic grains at 473 Ma, which is more compatible with the SHRIMP Arenig result. Interpretations of other MSID zircon ages from the Ordovician are also sensitive to choice of common Pb, and raise the likelihood that many multigrain ages might be too old owing to admixture with slightly older inherited zircon. A supposed 1–2% technical bias of SHRIMP 206Pb/238U ages relative to MSID is refuted.