Chloroplast genetic engineering overcomes concerns of gene containment, low levels of transgene expression, gene silencing, positional and pleiotropic effects or presence of vector sequences in transformed genomes. Several therapeutic proteins and agronomic traits have been highly expressed via the tobacco chloroplast genome but extending this concept to important crops has been a major challenge; lack of 100% homologous species-specific chloroplast transformation vectors containing suitable selectable markers, ability to regulate transgene expression in developing plastids and inadequate tissue culture systems via somatic embryogenesis are major challenges. We employed a ‘Double Gene/Single Selection (DGSS)’ plastid transformation vector that harbors two selectable marker genes (aphA-6 and nptII) to detoxify the same antibiotic by two enzymes, irrespective of the type of tissues or plastids; by combining this with an efficient regeneration system via somatic embryogenesis, cotton plastid transformation was achieved for the first time. The DGSS transformation vector is at least 8-fold (1 event/2.4 bombarded plates) more efficient than ‘Single Gene/Single Selection (SGSS)’ vector (aphA-6; 1 event per 20 bombarded plates). Chloroplast transgenic lines were fertile, flowered and set seeds similar to untransformed plants. Transgenes stably integrated into the cotton chloroplast genome were maternally inherited and were not transmitted via pollen when out-crossed with untransformed female plants. Cotton is one of the most important genetically modified crops ($ 120 billion US annual economy). Successful transformation of the chloroplast genome should address concerns about transgene escape, insects developing resistance, inadequate insect control and promote public acceptance of genetically modified cotton.