We show that Solar System tests can place very strong constraints on K-mouflage models of gravity , which are coupled scalar field models with nontrivial kinetic terms that screen the fifth force in regions of large gravitational acceleration .
In particular , the bounds on the anomalous perihelion of the Moon imposes stringent restrictions on the K-mouflage Lagrangian density , which can be met when the contributions of higher-order operators in the static regime are sufficiently small .
The bound on the rate of change of the gravitational strength in the Solar System constrains the coupling strength \beta to be smaller than 0.1 .
These two bounds impose tighter constraints than the results from the Cassini satellite and Big Bang Nucleosynthesis .
Despite the Solar System restrictions , we show that it is possible to construct viable models with interesting cosmological predictions .
In particular , relative to \Lambda -CDM , such models predict percent-level deviations for the clustering of matter and the number density of dark matter haloes .
This makes these models predictive and testable by forthcoming observational missions .