The emergence of lipid membranes and embedded proteins was essential for the evolution of cells. Translocon complexes mediate co-translational recruitment and membrane insertion of nascent proteins, but they already contain membrane-integral proteins. Therefore, a simpler mechanism must exist, enabling spontaneous membrane integration while preventing aggregation of un-chaperoned protein in aqueous phase. Here, we used giant unilamellar vesicles encapsulating a minimal protein synthesis kit to systematically interrogate the requirements for membrane insertion of proteorhodopsin (PR) – a structurally ubiquitous membrane protein. We show that the N-terminal hydrophobic domain of PR is both necessary and sufficient for co-translational recruitment of ribosomes to the membrane and subsequent membrane insertion of PR. Insertion of N-terminally truncated PR was restored by artificially attaching ribosomes to the membrane. Our findings offer a self-sufficient protein-inherent mechanism as a possible explanation for effective membrane protein biogenesis in a ‘pre-translocon’ era, and they offer new opportunities for generating artificial cells.