While stromal interactions are essential in cancer adaptation to hormonal therapies, the effects of bone stroma and androgen deprivation on cancer progression in bone are poorly understood. Here, we tissue-engineered and validated an in vitro microtissue model of osteoblastic bone metastases, and used it to study the effects of androgen deprivation in this microenvironment. The model was established by culturing primary human osteoprogenitor cells on melt electrowritten polymer scaffolds, leading to a mineralized osteoblast-derived microtissue containing, in a 3D setting, viable osteoblastic cells, osteocytic cells, and appropriate expression of osteoblast/osteocyte-derived mRNA and proteins, and mineral content. Direct co-culture of androgen receptor-dependent/independent cell lines (LNCaP, C4-2B, and PC3) led cancer cells to display functional and molecular features as observed in vivo. Co-cultured cancer cells showed increased affinity to the microtissues, as a function of their bone metastatic potential. Co-cultures led to alkaline phosphatase and collagen-I upregulation and sclerostin downregulation, consistent with the clinical marker profile of osteoblastic bone metastases. LNCaP showed a significant adaptive response under androgen deprivation in the microtissues, with the notable appearance of neuroendocrine transdifferentiation features and increased expression of related markers (dopa decarboxylase, enolase 2). Androgen deprivation affected the biology of the metastatic microenvironment with stronger upregulation of androgen receptor, alkaline phosphatase, and dopa decarboxylase, as seen in the transition towards resistance. The unique microtissues engineered here represent a substantial asset to determine the involvement of the human bone microenvironment in prostate cancer progression and response to a therapeutic context in this microenvironment.