Soil samples from grassland plots in the Hase floodplain near Osnabrück (NW Germany) were analyzed using visible and near‐infrared laboratory spectroscopy (VNIRS) by means of an ASD FieldSpec II Pro FR‐ spectroradiometer (spectral range 0.4–2.5 µm), paralleled by atomic‐absorption spectrometry (AAS), automated combustion for Corg quantification, and texture analysis. By AAS, contents of Cu, Zn, and Pb were found to be clearly elevated which is due to industrial effluents into the river in the 19th and 20th century. As these heavy metals (HMs) cannot be assessed directly by VNIRS, it was one major task to clarify whether they can be quantified indirectly using intercorrelations with spectrally active soil components. A second goal was to identify the specific spectroscopic predictive mechanism which may also be applicable to assess trace‐HM contents of other soil samples similar to those investigated here. For the latter, Corg was found to be most indicative, whereas the binding of the metals to other constituents (Fe oxides, clay) was not utilizable in the spectroscopic approach. The measured spectra were subjected to a multiplicative scatter correction and afterwards used to establish partial least‐squares regression (PLSR) models to estimate the contents of the different soil constituents. For Corg, very reliable estimates were obtained for both calibration and validation samples (in the validation, r² amounted to 0.90 and the percentage root mean square error [RMSE] was equal to 30.6%). Estimation accuracies obtained by PLSR for the trace HMs were considerably lower (r² between 0.56 and 0.71, percentage RMSE > 50%), which can be traced back to moderate correlations with Corg as main spectral determinant. According to these results, VNIRS can be applied as rapid and precise screening method for Corg to complement traditional analytical methods and to be used efficiently for a large number of samples. The method of VNIRS may also be applied for an indirect estimation of Corg‐associated metals to address their spatial variability, for example. This issue appears to be of high importance for digital soil mapping by imaging spectroscopy on a local to medium spatial scale.