Abstract Simon et al. (2005) reported low Ti 3+/Ti 4+ values in Ti-rich pyroxenes in the Wark–Lovering rim (WL) of a Leoville CAI (144A) as compared to the interior of the inclusion. These electron microprobe analyses were interpreted as evidence that the growth of the WL rim is the manifestation of an evolution to a more oxidizing environment. Further work by Simon et al. (2007) used XANES analyses to argue for higher Ti 3+ abundances and interpreted the data of Simon et al. (2005) as the result of X-ray contamination by neighboring phases, specifically spinel. Late-stage alteration was also included as a possible explanation. To investigate further the oxidation state of Ti in WL rims, we re-analyzed Leoville 144A to obtain a more complete data set of Ti 3+/Ti 4+ values in the Wark–Lovering rims. We conducted experiments on spinel-mixing to determine whether this was a plausible explanation for the observed paucity of Ti 3+ in WL rims. While we found a wider range of Ti 3+/Ti 4+ in these WL rim data than in our original study, our new data show that the original conclusion that rims are lower in Ti 3+/Ti 4+ than interiors remains valid. We conclusively rule out spinel-mixing as an explanation for our data, and we see no clear inconsistency between our electron microprobe data and the XANES data. The WL rim of CAI Ef3 was also analyzed by EMPA and compared to the results of Leoville 144A. To predict compositional consequences of this hypothesis, we constructed a reaction space between Ti-rich pyroxene in the WL rim, perovskite, Mg (g), Ca (g), O 2(g), and SiO (g). We find the oxidation of Ti 3+, coupled with Ti loss via perovskite formation, explains many features of WL rim EMPA analyses. We maintain that the WL rim pyroxenes are compositionally distinct from those in the interior, and are evidence of a more oxidizing environment during WL rim formation.