The development of plasma biomarkers has proven to be more challenging than initially anticipated. Many studies have reported lists of candidate proteins rather than validated candidate markers with an assigned performance to a specific clinical objective. Biomarker research necessitates a clear rational framework with requirements on a multitude of levels. On the technological front, the platform needs to be effective to detect low abundant plasma proteins and be able to measure them in a high throughput manner over a large amount of samples reproducibly. At a conceptual level, the choice of the technological platform and available samples should be part of an overall clinical study design that depends on a joint effort between basic and clinical research. Solutions to these needs are likely to facilitate more feasible studies. Targeted proteomic workflows based on SRM mass spectrometry show the potential of fast verification of biomarker candidates in plasma and thereby closing the gap between discovery and validation in the biomarker development pipeline. Biological samples need to be carefully chosen based on well-established guidelines either for candidate discovery in the form of disease models with optimal fidelity to human disease or for candidate evaluation as well-designed and annotated clinical cohort groups. Most importantly, they should be representative of the target population and directly address the investigated clinical question. A conceptual structure of a biomarker study can be provided in the form of several sequential phases, each having clear objectives and predefined goals. Furthermore, guidelines for reporting the outcome of biomarker studies are critical to adequately assess the quality of the research, interpretation and generalization of the results. By being attentive to and applying these considerations, biomarker research should become more efficient and lead to directly translatable biomarker candidates into clinical evaluation.