Nontargeted analysis of water samples using liquid chromatography combined with high-resolution mass spectrometers is an emerging approach for surface water monitoring and evaluation of water treatment processes. In this study, sample preconcentration via direct, large volume injection with 500 mu L and 1000 mu L injection volumes was compared to SPE regarding analytical performance parameters in targeted and nontargeted workflows. In targeted analysis, the methods were evaluated in terms of LOD and intrabatch precision of the selected compounds, whereas in nontargeted analysis, the number of detected unknown compounds, the method's intra-batch precision, and the retention time versus molecular mass pattern of the detected unknowns were evaluated. In addition, a novel intensity drift correction method was developed that is not based on quality control samples and makes use of the signals obtained for continuously infused reference compounds, which are conventionally utilized for online mass drift correction. It could be demonstrated that the new correction method significantly reduced the bias introduced by instrumental drift and is important for the reliable intercomparison of different nontargeted methods. Intercomparison of results showed that the 1000 mu L large volume injection method revealed the best performance in terms of precision under repeatability conditions of measurement as well as lower LODs for targeted compound analysis. In nontargeted analysis, the SPE method detected a higher number of unknown compounds but exhibited also a higher uncertainty of measurement caused by matrix effects.