Sodium gluconate is known to be a strong set retarder of cement. Many studies have been devoted to the interactions between cement and gluconate at early ages but the true mechanisms of this retardation is not yet elucidated. The most important phase in cement is alite which is a tricalcium silicate containing impurities. Tricalcium silicate hydration leads to the nucleation and growth of calcium silicate hydrates which are responsible for setting and hardening of cement paste. This work is devoted to the influence of sodium gluconate on the early hydration, period during which calcium silicate hydrates should nucleate and begin to grow on anhydrous grains. Previous studies have shown that the delay before acceleration of hydration, i.e. calcium silicate hydrates growth, mainly depends on this period. The systematic follow up of ionic concentrations of diluted system suspensions during hydration of pure C3S and also cement enable to study the C-S-H nucleation-growth period. During hydration of pure C3S, a complex calcium-gluconate adsorbs on first C-S-H nuclei and prevents them from growing. During hydration of cement, there is mainly formation of a complex aluminium-gluconate, aluminium coming from aluminate phases dissolution. This complex also prevents first precipitated C-S-H nuclei from growing.