ERK and p38 MAP kinases, acting through the downstream mitogen- and stress-activated kinase 1/2 (MSK1/2), elicit histone H3 phosphorylation on a subfraction of nucleosomes--including those at Fos and Jun--concomitant with gene induction. S10 and S28 on the H3 tail have both been shown to be phospho-acceptors in vivo. Both phospho-epitopes appear with similar time-courses and both occur on H3 tails that are highly sensitive to TSA-induced hyperacetylation, similarities which might suggest that MSK1/2 phosphorylates both sites on the same H3 tails. Indeed, on recombinant histone octamers in vitro, MSK1 efficiently phosphorylates both sites on the same H3 tail. However, sequential immunoprecipitation studies show that antibodies against phosphorylated S10-H3 recover virtually all this epitope without depletion of phosphorylated S28-H3, and vice versa, indicating that the two phospho-epitopes are not located on the same H3 tail in vivo. Confocal immunocytochemistry confirms the clear physical separation of the two phospho-epitopes in the intact mouse nucleus. Finally, we used transfection-based experiments to test models that might explain such differential targeting. Overexpression and delocalisation of MSK1 does not result in the breakdown of targeting in vivo despite the fact that the ectopic kinase is fully activated by external stimuli. These studies reveal a remarkable level of targeting of S10 and S28 phosphorylation to distinct H3 tails within chromatin in the interphase mouse nucleus. Possible models for such exquisite targeting are discussed.