Key words∂-Neumann problem,∂ b -system, q-pseudoconvex/concave manifolds. MSC (2010) 32F10, 32F20, 32N15, 32T25We study ellipticity in a weak sense, such as fractional or logarithmic, of the system ∂ b ,∂ * b tangential to a hypersurface or a generic higher codimensional submanifold M ⊂ C n . The geometric setting which assures the estimates is the q-pseudoconvexity/concavity of M in addition to the existence of a suitable family of weights in a strip or a tube around M . The basic estimates for the∂-Neumann problem on q-pseudoconvex/concave domains is related to the classical work by Shaw [17] and more recent by Zampieri [19]. The method of the weights is due to Catlin [3] and the relation between the tangential and the ambient∂ system on pseudoconvex domains is inspired to Kohn [14]. Both these techniques are adapted here to a general Levi signature.Let M be a real smooth hypersurface of C n defined by r = 0 with ∂r = 0. We denote by D = D + and D = D − the two sides of M in a neighborhood V of a point z o ∈ M ; we assume r < 0 in D + and r > 0 in D − . Let ω 1 , . . . , ω n be an orthonormal basis of (1, 0) forms in a neighborhood of z o with ω n = ∂r, and let ∂ ω1 , . . . , ∂ ωn be the dual basis of (1, 0) vector fields. For 0 ≤ k ≤ n, we write a general k-form u on V aswhere denotes summation restricted to ordered multiindices J = {j 1 , . . . , j k } and whereω J =ω j1 ∧ · · · ∧ ω j k . When the multiindex is no more ordered, it is understood that the coefficient u J is antisymmetric with respect to J; in particular, if J decomposes into jK, then u j K = sign J j K u J . We define a scalar product and a norm by u, u = |u| 2 = |J |=k |u J | 2 ; this definition is independent of the choice of the orthonormal basis ω 1 , . . . , ω n . The coefficients of our forms are taken in various spaces such asand the corresponding spaces of k-forms are denoted by C ∞ (D ∩ V ) k and so on. All our discussion is local; sometimes, we omit this specification. Though our a priori estimates are proved over smooth forms, they are stated in Hilbert norms. Thus, let u H 0 or u 0 be the H 0 = L 2 norm and, for a real function ϕ, let the weighted L 2 -norm be defined bywhere dv is the element of volume in C n .