Background In the era of precision oncology, NGS-based detection of tumor-specific oncogenic alterations has become important to select patients for targeted therapies. Whole genome sequencing (WGS) offers the most comprehensive analysis of genetic alterations. Yet, tumor WGS has not been used widely for diagnostic purposes due to the reduced quality and quantity of DNA from FFPE material. Here, we present a WGS protocol for sequencing and analysis of tumor normal pairs from FFPE material and blood with an input of 100 ng of DNA. Methods WGS libraries were generated from peripheral blood DNA and FFPE tumor DNA using NEBNext UltraShear FFPE DNA Library Prep Kit (New England Biolabs) with 3 and 5 cycles of PCR amplification, respectively. Tumor-normal DNA sample pairs of 15 patients were sequenced by WGS and by whole exome sequencing (WES). WES libraries were processed using a Twist Custom Exome enrichment kit (Twist Bioscience). For three patients, tumor WGS libraries were prepared from fresh frozen (FF) and from FFPE material. Libraries were sequenced in paired-end mode on an Illumina platform, followed by data analysis using an in-house bioinformatics pipeline (megSAP). Results WGS was able to recapitulate the results from diagnostic WES in 15 pairs of tumor-normal DNA samples. Mean coverage was 171 for tumor- and 53 for normal-DNA. Single nucleotide variants were detected with a sensitivity of 95 % and precision of 99 % at a somatic allele frequency of 10 %. Direct comparison of WGS results of three samples where both FF and FFPE material was analyzed showed a superiority of FF material in the detection of structural variants (SVs) and of somatic variants with low tumor fraction. The diagnostic yield of somatic alterations related to targeted therapies was higher in WGS analysis in 7 out of 15 samples compared to WES analysis because of an enhanced detection of SVs. Conclusions The developed workflow for tumor WGS from FFPE material can be used as a diagnostic test in the context of molecular tumor board-guided targeted treatment. The presented protocol allows the processing of low-quality DNA samples. Despite of higher sequencing costs, the diagnostic power of WGS is superior to WES.