Problem. To evaluate the total concentration ( с Σ ) of hydrocarbons (HBs) in natural and waste waters it is necessary to extract all HBs, separate them from non-HB substances and measure the generalized signal in the certain IR spectral region. Then one may to determine total HB index ( с *) recalculated to the certain standard substance. The error δ с = ( с * - с Σ ) / с Σ has to depend on the method of signal measuring but no systematic investigations of this problem were early conducted. Experiment. The model mixtures of alkanes, cycloalkanes and alkilarenes in CCl 4 were prepared. Two methods of generalized signal measuring were used. Namely, one-wave IR-spectrometry (FT-801 Fourier-spectrometer) and IR-photometry (concentratometer ICN-025 with interference filters). The total HB index was determined for each sample, the same standard (Simard mixture) was used for all determinations. The both estimations for each sample were compared. Results. For both methods resulting δ с –values are repeatable and significant. The measurement of the average absorbance near 3.3-3.5 mcm by IKN-025 concentratometer leads to more accurate (±20-30 % rel.) estimations of the total hydrocarbon content than the measurement of the monochromatic absorbance near 3.42 mcm by the FT-801 Fourier-spectrometer (errors to the point of 80 % rel.), with other equal conditions. The one-wave IR spectrometry in comparison with IR-photometry has the only benefit - low LODs of hydrocarbons. Discussion. Irrespective of the way to measure the signal, the error δ с arises mainly because the real mismatch of standard and model mixtures in composition exists. The other factor is the inequality of the absorption coefficients for different HBs. This hypothesis is confirmed, as we are able to predict δ с –values taking into the consideration the absorption coefficients and the composition of mixtures to investigate. The leveling of these absorption coefficients for various HBs leads to more reliable estimations for с Σ . The probable mechanism for absorption coefficients leveling is proposed on the basis of spectral data. Evidently, the multiwave IR-spectrometry and integral intensities measuring have to show the best accuracy of hydrocarbon content estimations and the low LODs simultaneously. In the ”Conclusion” the fundamental and applied significances of the investigation are characterized. Key words: aqueous analysis, total indices, petroleum products, one wave IR spectrometry, IR photometry, concentratometers, errors of the analysis, limits of detection. (Russian) DOI: http://dx.doi.org/10.15826/analitika.2014.18.1.010 M.A. Fedorova, S.V. Usova, V.I. Vershinin Dostoevsky Omsk State University , Omsk, Russian Federation