Cisplatin is a classic anticancer drug widely used as a reference drug to test new metal complex drug candidates. We found an unexpected diversity in cisplatin-related cytotoxicity values, expressed as IC50 (the half-maximal inhibitory concentration) in tumour cell lines, such as MCF-7, HepG2 and HeLa. We reviewed the data published from 2018 to 2022. A total of 41 articles based on 56 in vitro experiments met our eligibility criteria. Using a meta-analysis based on a random effect model, we evaluated the cytotoxicity of cisplatin (IC50) after 48- or 72-h cell exposure. We found large differences between studies using a particular cell line. According to the random effect model, the 95% confidence intervals for IC50 were extremely wide. The heterogeneity of cisplatin IC50, as measured by the I2 index for all cancer cell lines, was over 99.7% at culture times of 48 or 72 h. Therefore, the variability between studies is due to experimental heterogeneity rather than chance. Despite the higher IC50 values after 48 h than after 72 h, the heterogeneity between the two culture periods did not differ significantly. This indicates that the duration of cultivation is not the main cause of heterogeneity. Therefore, the available data is diverse and not useful as a reference. We discuss possible reasons for the IC50 heterogeneity and advise researchers to conduct preliminary testing before starting experiments and not to solely rely on the published data. We hope that this systematic meta-analysis will provide valuable information for researchers searching for new cancer drugs using cisplatin as a reference drug.