The development of an axillary bud of white clover to form a branch depends on (1) the bud being viable, vegetative and non-dormant, and (2) suitable conditions for outgrowth of the bud. Foraging theory emphasises the second of these requirements. Glasshouse studies with white clover rarely result in a loss of bud viability. In contrast, in field populations over 50% of the buds reaching the stage of maturity when branching can occur are not in a viable, vegetative, non-dormant condition. We examined whether non-viability could be induced in a glasshouse experiment by applying treatments in a factorial design. The factors were: defoliation, phosphorus supply, soil moisture status, simulated treading and grass competition. In addition, we measured the effects of the treatments on the outgrowth of viable buds in order to assess whether the same factors were determining viability and outgrowth. Defoliation significantly reduced bud viability (by 44%) but no other factors, either singly or in combination, had a significant effect. A greater variety of factors and combinations of factors influenced bud outgrowth; these were defoliation, phosphorus status and interactions involving phosphorus and grass; defoliation, phosphorus and soil moisture; and soil moisture, grass and treading. For white clover it is relevant to include the state of the axillary meristem in any model of foraging. (C) 1996 Annals of Botany Company