One-cell mouse embryos of the Balb/c strain normally divide at 18··5 h p.c. (post conception), but they suffer an extremely long G2arrest when irradiated with 2 Gy X-rays 8 h p.c. at the early pronuclear stage. This could be an indirect effect of radiation on tyrosine dephosphorylation of the p34cdc2subunit of a maturation or mitosis promoting factor (MPF), which normally occurs at the end of G2. This, in turn, would maintain MPF in an inactivated form and block entry into mitosis. Preliminary studies were undertaken at the morphological level to assess indirectly the validity of this hypothesis. For this purpose, irradiated and control embryos were exposed to different compounds, which are known to interfere, directly or indirectly, with the state of phosphorylation/dephosphorylation of p34cdc2. Caffeine (CAF; 2 mm) did not affect the time of first division of control embryos, but it completely suppressed the radiation-induced G2arrest of embryos exposed to this compound from 17 h p.c., i.e. 1··5 h before the normal time of first cleavage. Under the same conditions, okadaic acid (OA; 3 μμm), a specific inhibitor of phosphatases I and IIA, induced a rapid pronuclear membrane breakdown and a block of all control and irradiated embryos at metaphase. Genistein (GEN; 92 or 185 μμm). a potent inhibitor of tyrosine kinases, increased the radiation-induced G2arrest and even induced a dose-dependent G2arrest in the control embryos. Embryos were exposed at different times following irradiation to a mixture of either CAF (2 or 5 mm) or OA (3 or 10 μμm), and cycloheximide (CH; 5 μμg/ml), a potent protein synthesis inhibitor. Reversion of G2-arrest by CAF was still seen in embryos exposed to CAF + CH from 17 h p.c. However, the proportion of irradiated embryos eventually able to cleave was lower than that obtained under the conditions of exposure to CAF alone. Embryos exposed to CAF + CH before 17 h p.c. were not able to cleave, regardless of the concentration of CAF used. Nuclear envelope breakdown still occurred in 100% control and irradiated embryos, following exposure to 3 μμm OA + CH from 10 h p.c., or to 10 μμm OA + CH from 8··5 p.c. Embryos, which entered premature division, showed in their pronuclei the presence of PCC (premature chromosome condensation) in either S or G2, depending upon when they had been exposed to OA + CH. The mechanism of G2arrest in embryos exposed to GEN in late G2remains to be defined, as does the mechanism of the complete reversal of G2arrest induced by CAF. Our results suggest that CAF and OA act in different ways to suppress the radiation-induced G2arrest. Thus, OA could induce direct dephosphorylation of p34cdc2via a type 2A phosphatase, although the levels of cyclin that are normally required for this process have not yet been reached. CAF apparently acts in a much less direct way, possibly inducing MPF activation by affecting upstream proteins which regulate the process. The dependence of CAF on cyclin and/or other protein(s) for such action is also clearly more pronounced. As with GEN, the elucidation of the exact mechanisms of action of CAF and OA will require detailed biochemical studies.