A blank peak effect in FIA systems often found for samples containing an excess of solute other than the analyte has been studied. Emphasis has been given to trace level determinations and a simple approach for elimination of this effect is presented. The present approach is based on the use of a large sample injection volume, which results in a portion of the sample plug being undiluted with carrier and hence prevents the formation of refractive index gradients in the undiluted portion. The quantitative performance of this approach was demonstrated with the determination of iron in the presence of 0.20 M sodium chloride and cobalt in the presence of 0.10 M glucose. No significant difference in the accuracy, precision, and limit of detection was observed between samples for iron both in the presence and absence of sodium chloride and for cobalt in the presence and absence of glucose. Despite using a large sample volume (0.89 ml for a 400-cm sample loop), the sample throughput was about 25/hr.