We analyse the metrological aspect related to systems devoted to the simultaneous measurements of (one point) velocity and scalar concentration statistics. We consider systems based on the coupling of an anemometer (laser Doppler or hot wire) with a flame ionization detector. In particular, we focus on the estimate of the cross-correlation velocity-concentration. Key aspects in the reliability of such measurements are related to the identification of an optimal distance between the two probes, the time lag with which the velocity and concentration signals are acquired, and, eventually, the influence of seeding particles (when using a laser Doppler anemometer) on the concentration measurements. We investigate these aspects by performing wind-tunnel experiments on the scalar dispersion downwind a ground-level source in a turbulent boundary layer. This analysis allowed us to identify a time delay and a convenient distance between the signals provided by the anemometer and the concentration detector. Results also show that the seeding used for the laser Doppler anemometer measurements modifies the cut-off frequency in the concentration spectra and induces a slightly larger uncertainty on the statistics of scalar field. The results show the reliability of both systems in estimating (horizontal and vertical) turbulent mass fluxes.