An inverse-problem approach is proposed in order to improve the accuracy of tracer flow measurements. Such an approach is useful when geometrical pipe configurations are drastic (e.g. for pipes with singularities such as bends). In this approach, the concentration profile is reconstructed according to a parametric model. The parameter-identification process requires only observations (observed signals), i.e. no additional information is introduced. We compare this approach with a classical numerical technique. For that purpose, velocity and concentration profiles in circular pipes are worked out with a computational fluid dynamics code in various cases of singularity. These simulations lead to the reference data required for comparison. The inverse-problem approach led to better results than the classical one.