We show that the maximally supersymmetric vacua of minimal d 5 N 1 sugra remain maximally supersymmetric solutions when taking into account higher order corrections.The question of whether a given supergravity solution is a consistent background for string propagation to all orders in perturbation theory is an interesting though hard one. It has of course long been known that pp-waves with flat transverse space provide such backgrounds since all the scalar invariants vanish identically -recently the class of exact sugra solutions with vanishing scalar invariants was investigated in [1]-but they are exceptional.A class of sugra solutions for which a proof of all-order consistency is highly desirable are the maximally supersymmetric solutions, and for most of them an answer is known: in Ref.[2], Kallosh and Rajaraman, by making use of superspace methods, showed the all-order consistency of aDS 2 S 2 in minimal N 2 d 4 sugra, of aDS 5 S 5 in type IIB, and also of aDS 4 S 7 and aDS 7 S 4 in Mtheory. Since the associated Minkowski and KowalskiGlikmann solutions [3][4][5] are pp-waves, we must conclude that all the maximally supersymmetric solutions in minimal N 2 d 4, type IIB-sugra or M-sugra are allorder consistent.The way Kallosh and Rajaraman attacked the problem leans heavily on the fact that the Riemann tensor and the field strengths are covariantly constant with respect to (w.r.t.) the Levi-Cività connection. This covariantly constancy, however, is due to the fact that the solutions describe symmetric spacetimes, G=H, with G-invariant field strengths. Interestingly, the vast majority of maximally supersymmetric solutions are described by symmetric spaces, and one can envisage similar arguments to apply for their consistency.The strange ducks in the pond are the maximally supersymmetric solutions in d 5 sugra [6 -10]. Solutions such as the near-horizon-BMPV solution or the Gödel space, are not symmetric [11]: rather, they describe homogeneous, naturally reductive spacetimes with compatible fluxes. As such, there is a metric compatible connection that parallelizes the Riemann tensor and the field strengths, but it is not the Levi-Cività one. We then should ask ourselves the question whether the maximally supersymmetric solutions of d 5 N 1 sugra are all-order exact or not, and how to attack the problem. The answer to the last question lies in the way one would construct, and indeed constructs, higher order sugra actions in lower dimensions, be they string inspired or not.Partial results have recently been obtained in Refs. [12 -14] and these works instigated the current investigations: the symmetric solution aDS 2 S 3 was shown to be maximally supersymmetric in Ref. [14] and aDS 3 S 2 in Ref.[13] in a theory with TrA^2 R corrections. This theory was constructed in [12] and they also derived the conditions for the existence of a maximally supersymmetric aDS 5 .Dealing with on-shell sypersymmetry in systems with higher orders is quite cumbersome: since it is on-shell, the supersymmetry transformations ''need to kn...