The replacement of cement by mineral admixtures in concrete has been of increasing interest in the construction industry. Nevertheless, several of the potential replacements, such as fly ash class F, lower the compressive strength of concrete at early age. This project investigates the use of nanosilica to compensate for such loss of compressive strength. A statistical experimental design involving mixtures of Portland cement, fly ash and nanosilica, in addition to water/binder ratio as an external factor, is proposed to study their combined effect on the compressive strength of concrete. This design allows estimating a cubic regression model that properly accounts for the effects of the mixture components within a constrained experimental region. The range of each factor was selected according to levels normally used in the industry. Finally, an optimisation strategy permits to recommend the use of nanosilica when high percentages of cement replacement by fly ash are present.