Although $\Lambda$CDM fits a broad set of data, the nature of dark energy and the cosmological constant problem motivate tests of modified gravity, backreaction, and large-scale inhomogeneity. Any alternative must reproduce the successes of standard cosmology across many observables.

Common approaches include phenomenological functions such as $\mu(k,a)$ and $\eta(k,a)$ that modify the Poisson equation and gravitational slip, or specific scalar-tensor theories and $f(R)$ gravity. These frameworks are compared through growth and lensing observables.

A subtle research topic is whether the averaging of inhomogeneous spacetime can produce effective acceleration. Most analyses find that realistic backreaction effects are too small to replace dark energy, but the problem remains conceptually important in relativistic cosmology.

Many modified-gravity theories invoke screening, such as chameleon, symmetron, or Vainshtein mechanisms, to recover general relativity in high-density environments while allowing departures on cosmic scales.