Can enzymatically dispersed testicular cells from adult men reassemble into seminiferous cord-like structures in vitro? Adult human testicular somatic cells reassembled into testicular cord-like structures via dynamic interactions of Sertoli and peritubular cells. In vitro approaches using dispersed single cell suspensions of human testes to generate seminiferous tubule structures and to initiate their functionality have as yet shown only limited success. Testes from 15 adult gender dysphoria patients (mean ± standard deviation age 35 ± 9.3 years) showing spermatogonial arrest became available for this study after sex-reassignment surgery. In vitro primary testicular somatic cell cultures were generated to explore the self-organizing ability of testicular somatic cells to form testis cords over a 2-week period. Morphological phenotype, protein marker expression and temporal dynamics of cell reassembly were analyzed. Cell suspensions obtained by two-step enzymatic digestion were plated onto glass coverslips in 24-well plates. To obtain adherent somatic cells, the supernatant was discarded on Day 2. The culture of the attached cell population was continued. Reassembly into cord-like structures was analyzed daily by microscopic observations. Endpoints were qualitative changes in morphology. Cell types were characterized by phase-contrast microscopy and immunohistochemistry. Dynamics of cord formation were recorded by time-lapse microscopy. Primary adult human testicular cells underwent sequential morphological changes including compaction and reaggregation resulting in round or elongated cord-like structures. Time-lapse video recordings within the first 4 days of culture revealed highly dynamic processes of migration and coalescence of reaggregated cells. The cellular movements were mediated by peritubular cells. Immunohistochemical analysis showed that both SRY-related high mobility box 9-positive Sertoli and α-smooth muscle actin-positive peritubular myoid cells interacted and contributed to cord-like structure formation. Not applicable. Owing to scarcity of normal human testicular tissue, testes from gender dysphoria patients were used in the study. The regressed status might influence the experimental responses of primary cells. We observed basic morphological features resembling in vivo testicular cords, however, the proof of functionality (e.g. support of germ cells) will need further studies. The proposed in vitro culture system may open opportunities for examination of testicular cell interactions during testicular tubulogenesis. Further refinement of our approach may enable initiation of ex vivo spermatogenesis. The work was supported by EU-FP7-PEOPLE-2013-ITN 603568: 'Growsperm'. No conflict of interests is declared.