In this paper, an analysis of the conditioning of methods solving the inverse heat conduction problem is shown. A method is developed that connects the stability of the inverse solution, measured by the condition number, to the physical properties of the problem. The approach is based on norm and eigenvalue analysis of the matrix representation of the discretized inverse heat conduction problem. This formula can be applied to methods that are based on the utilization of the heat kernel. It can be used in order to determine the optimal setup for surface heat flux measurements.regularized by sequential function estimation method A VC = A regularized by van Cittert scheme a ij = element of A c = specific heat capacity d = distance to the nearest surface, mm g = arbitrary function h = fundamental solution, K · m 2 ∕J h a = regularized time-discrete fundamental solution, K · m 2 ∕J h p = numerical time discrete fundamental solution, K · m 2 ∕J k = thermal conductivity, W∕m · K _ q = surface heat flux density, W∕m 2 T = temperature, K t = time, s t f = regularization time, s t p= penetration time, s x, y, z = spatial variables, mm α = thermal diffusivity, m 2 ∕s δA = modeling error δ _ q = surface heat flux error δT = temperature measurement error θ = time-step size, s θ 0 = regularization time shift, s κ = condition number λ = eigenvalue ξ, τ = integration variables ρ = density, kg∕m 3 χ = dimensionless time-step size; α · θ∕d 2 k · k F = Frobenius norm k · k 1 , k · k ∞ = operator norms k · k 2 = spectral norm