Prostate cancer is caused by genomic aberrations in normal epithelial cells, however clinical translation of findings from analyses of cancer cells alone has been very limited. A deeper understanding of the tumour microenvironment is needed to identify the key drivers of disease progression and reveal novel therapeutic opportunities. We established the most extensive cell-type specific transcriptomic landscape of prostate tumour microenvironment to date; profiling 52 samples from 13 patients, that were enriched for one of four key cell types. For enhanced transcriptional analyses, we developed a novel Bayesian method, TABI, which avoids a priori patient risk stratification, and maps transcriptional alteration events to specific risk scores. As result, we produced a map of concurrent cell-type transcriptional alterations, whose clinical relevance was tested on an extended independent prostate cancer dataset. An important role of monocytes and macrophages in prostate cancer progression and disease recurrence was uncovered, supported by both our transcriptional landscape findings and by differential tissue composition analyses.