Data from vertical net hauls and from a submersible optical plankton counter (OPC) were compared in terms of biomass and slope of the normalized biomass-size spectra (NB-SS), a proxy for the size structure of the community. The relationship between the estimates of biomass in the overlapping range sampled by both methods (0.2-2 mill equivalent spherical diameter (ESD)) was linear and not significantly different to I (B-OPC:B-NET = 1.1 +/- 0.1, r(2) = 0.84). However, the relationship varied depending oil the size fraction considered; the ratio B-OPC:B-NET was 0.10 +/- 0.04 (r(2) = 0.30), 0.8 +/- 0.2 (r(2) = 0.66), and 2.4 +/- 0.5 (r(2) = 0.64) for the size fractions 0.2-0.5, 0.5-1. and 1-2 mm ESD, respectively. The discrepancies between methods were presumably due to the combined effect of the limitations of the instrument in the lower detection limit and the errors in the volume of water sampled by the net for the smallest size fraction, and to net avoidance enhanced by clogging for the largest size fraction. The agreement between methods improved when the NB-SS of the different data sets were compared. The slope (b = -1.1) and the intercept (a = 14.6) of the NB-SS integrated across stations were not significantly different (Student's t-test) for the linear model fitted to net samples, OPC data, or pooled data from both methods. Station by station, the slopes of the NB-SS from the net (b(NET)) and the OPC (b(OPC)) were not significantly different in 61% of the stations. This percentage increased to 78% when the comparison was limited to the mesoplankton size range. As an example of the applicability of the OPC, we showed the distribution of mesoplankton biomass and size structure along the NW and N Iberian Shelf during the winter-spring transition of 2002 and its relationship with the hydrographic scenario.