In this work, the possibilities for the direct determination of Cr in solid samples by means of graphite furnace atomic absorption spectrometry (SS-GFAAS) and by means of electrothermal vaporization-inductively coupled plasma mass spectrometry (SS-ETV-ICPMS) have been critically evaluated and compared, the goal always being to develop simple and fast methods, only relying on the use of aqueous standards for calibration. Four reference materials of very different nature (milk powder, lobster hepatopancreas, polyethylene and sewage sludge), covering a wide Cr concentration range (from 17.7 ng g 21 to 37.2 mg g 21 ), were selected for the study. The diverse volatility of their matrixes, compared with that of Cr, resulted in different degrees of difficulty for both techniques. SS-GFAAS offers a better limit of detection (2 ng g 21 ) than SS-ETV-ICPMS (30 ng g 21 ), mainly due to the occurrence of C-based polyatomic interferences with the latter technique. The samples that presented the least difficulties were those for which the matrix is mainly of an organic nature and can be easily removed during the pyrolysis step, allowing a selective atomization/vaporization of the analyte. Under these circumstances, external calibration against aqueous standards was feasible for SS-GFAAS in all cases. SS-ETV-ICPMS proved to be more sensitive to matrix effects, but the use of the palladium (added for its carrier properties) signal as internal standard also allowed the use of this straightforward calibration procedure. On the other hand, the analysis of sewage sludge (containing 85% of hardly volatile inorganic compounds) represented the most complicated situation for both techniques: for SS-GFAAS, the addition of Na 2 CO 3 to effect an in-situ microfusion of the sample during the pyrolysis step greatly helped to improve the situation, while for SS-ETV-ICPMS the single standard addition method was necessary. In all cases, a good agreement with the certified values was obtained. The precision ranged between 6 and 22% relative standard deviation, depending on the homogeneity and the sample mass used. These values could be further improved, if needed, by increasing the number of replicates per determination (usually five). Every determination required approximately 15-20 min, except when the single standard addition method was used (30-40 min). Therefore, it can be stated that both techniques show the capability of providing a direct determination of this analyte in a variety of samples, at very different concentration levels, allowing the development of fast and reliable solid sampling procedures.