Precision oncology aims to tailor cancer treatment to the biological characteristics of individual tumours. However, genomic profiling alone often fails to predict therapeutic response, highlighting the need for complementary functional approaches. Patient-derived tumour organoids (PDTOs) offer a promising platform to investigate tumour biology and therapeutic vulnerabilities. Nevertheless, their translational value depends on the reliability of their establishment, biological fidelity and experimental conditions. This thesis aimed to develop, optimize and critically evaluate a PDTO platform for translational cancer research. First, we established a multi-tumour organoid biobank at the University Hospital Antwerp and University of Antwerp, integrating standardized tissue collection, laboratory processing, quality control and clinical data management. Evaluation of 182 patient samples demonstrated substantial tumour-specific variability in establishment success, with long-term cultures obtained in 16.5% of cases. Challenges included limited tissue availability, variable growth capacity and non-malignant epithelial overgrowth, particularly in lung cancer cultures. These findings emphasize that successful biobanking requires both technical optimization and rigorous model validation. Second, we investigated the potential of PDTOs for functional drug discovery, focusing on strategies to enhance the efficacy of 5-fluorouracil (5-FU). Through iterative monotherapy, combination and pathway-focused screening, we identified inhibition of the replication-stress checkpoint, particularly the ATR–CHK1–WEE1–PKMYT1 axis, as a promising strategy to potentiate 5-FU activity. These findings illustrate the value of functional screening for identifying therapeutic vulnerabilities while emphasizing the importance of mechanistic validation. Third, we systematically investigated how culture medium influences organoid biology and pharmacological responses. Comparing conventional Wnt-based medium with OncoPro medium revealed substantial differences in transcriptional states and drug sensitivity despite preservation of major genomic alterations and tumour identity. OncoPro-grown organoids generally exhibited greater drug resistance and transcriptional profiles more closely resembling malignant epithelial states. These findings demonstrate that culture conditions are not merely technical parameters but active determinants of experimental outcomes, with important implications for reproducibility and interpretation of functional drug screening. Collectively, this thesis establishes a multi-tumour PDTO research infrastructure and demonstrates its potential for biological discovery and therapeutic exploration. Importantly, it also highlights that organoids are experimentally shaped models rather than unbiased replicas of patient tumours. Their future will require improved establishment efficiency, standardized experimental conditions, appropriate model complexity and prospective clinical validation. Understanding both the capabilities and limitations of PDTOs is essential to ensure their reliable application in translational cancer research.