We study the hidden symmetries of the fermionic sector of D = 11 supergravity, and the rôle of K(E 10 ) as a generalised 'R-symmetry'. We find a consistent model of a massless spinning particle on an E 10 /K(E 10 ) coset manifold whose dynamics can be mapped onto the fermionic and bosonic dynamics of D = 11 supergravity in the near space-like singularity limit. This E 10 -invariant superparticle dynamics might provide the basis of a new definition of M-theory, and might describe the 'de-emergence' of spacetime near a cosmological singularity. Eleven-dimensional supergravity (SUGRA 11 ) [1] is believed to be the low-energy limit of the elusive 'M-theory', which is, hopefully, a unified framework encompassing the various known string theories. Understanding the symmetries of SUGRA 11 is therefore important for reaching a satisfactory formulation of M-theory. Many years ago it was found that the toroidal dimensional reduction of SUGRA 11 to lower dimensions leads to the emergence of unexpected ('hidden') symmetry groups, notably E 7 in the reduction to four noncompactified spacetime dimensions [2], E 8 in the reduction to D = 3 [2-5], and the affine Kac-Moody group E 9 in the reduction to D = 2 [6,7]. It was also conjectured [8] that the hyperbolic Kac-Moody group E 10 might appear when reducing SUGRA 11 to only one (time-like) dimension.Recently, the consideration, à la Belinskii, Khalatnikov and Lifshitz [9], of the near space-like singularity limit 1 of generic inhomogeneous bosonic eleven-dimensional supergravity solutions has uncovered some striking evidence for the hidden rôle of E 10 [10,11]. Ref.[11] related the gradient expansion * Corresponding author.E-mail address: axel.kleinschmidt@aei.mpg.de (A. Kleinschmidt). 1 This limit can also be viewed as a small tension limit, α → ∞.(∂ x ∂ t ), which organises the near space-like singularity limit [12], to an algebraic expansion in the height of positive roots of E 10 . A main conjecture of [11] was the existence of a correspondence between the time evolution, around any given spatial point x, of the supergravity bosonic fields g (11) MN (t, x), A (11) MNP (t, x), together with their infinite towers of spatial gradients, on the one hand, and the dynamics of a structureless massless particle on the infinite-dimensional coset space E 10 /K(E 10 ) on the other hand. Here, K(E 10 ) is the maximal compact subgroup of E 10 . Further evidence for the rôle of the one-dimensional non-linear sigma model E 10 /K(E 10 ) in M-theory was provided in [13][14][15][16].An earlier and conceptually different proposal aiming at capturing hidden symmetries of M-theory, and based on the very-extended Kac-Moody group E 11 , was made in [17,18] and further developed in [19][20][21]. A proposal combining the ideas of [18] and [11] was put forward in [22][23][24].In this Letter, we extend the bosonic coset construction of [11] to the full supergravity theory by including fermionic variables; more specifically, we provide evidence for the existence of a correspondence between th...