The International Collaboration on Cancer Reporting (ICCR) is a project which issues datasets and guidelines for international standardisation of cancer reporting. This review summarises the required and recommended elements of the datasets for prostate core needle biopsies and transurethral resection (TURP) and enucleation specimens of the prostate. To obtain as much information as possible from needle biopsies there should be only one core in each specimen jar with the exception of saturation biopsies. The gross description of the specimens should include core lengths of needle biopsies and weight of resection specimens. The tumours should be classified according to the 4th World Health Organization (WHO) classification and graded both by Gleason scores and the grouping of these in International Society of Urological Pathology (ISUP) grades (Grade groups). Percent high-grade cancer is an optional component of the report. Tumour extent in needle biopsies should be reported both by number of cores positive for cancer and the linear extent measured in either millimetre or percent core involvement by tumour. In needle biopsies where low-grade cancer is discontinuous and seen in few cores, it is recommended that the tumour extent should be reported both by including and subtracting intervening benign tissue. For resection specimens, the percentage of the tissue area (or percentage of number of TURP chips) involved with cancer should be estimated. Extraprostatic extension should be reported when seen, while the reporting of perineural, seminal vesicle/ejaculatory duct and lymphovascular invasion is only recommended. Intraductal carcinoma of the prostate (IDC-P) should be reported when present, because of its strong link with aggressive cancer. The current recommendation is that the IDC-P component should not be graded. The structured and standardised reporting of prostate cancer contributes to safer and more efficient patient care and facilitates the compilation and understanding of multiparametric diagnostic and prognostic data.