For the past 75 years subcutaneous injections have been the only route of delivery of insulin therapy to diabetic patients. During this time, numerous attempts have been made to explore alternative routes for systemic insulin administration. However, thus far, no feasible other way of non-invasive insulin delivery has been developed. Dermal insulin application does not result in a reproducible and sufficient transfer of insulin across the highly efficient skin barrier. The dream of an "insulin tablet" has also not become a reality, the main problem being digestion and a lack of a specific peptide carrier system in the gut. Nasal insulin application was considered for a number of years as a potential method, because of the rapid absorption of insulin across the nasal mucosa. However, relative bioavailability was low and required use of absorption enhancers and more importantly, the metabolic effect lasted too short to be of clinical usefulness. To date the most promising alternative route of insulin administration, is the pulmonary delivery of insulin by inhalation which will likely lead to a practically usable system within the next few years. For maximal rate of absorption insulin must be applied deep into the lung, i.e., into the alveoli. A considerable number of inhalers (in combination with appropriate insulin formulations), which are ask to generate insulin particles with an appropriate size for pulmonary delivery, are currently in the clinical phase of development. The pharmaco dynamic effects of insulin formulations administered via the lung are comparable to, or even faster than, those of s.c. injected regular insulin or rapid-acting insulin analogues. The relative biopotency of inhaled insulin in most cases is approximately 10%, i.e., the dose of insulin administered must be 10-fold higher than with s.c. application. The published results of clinical trials thus far, indicate that metabolic control is comparable to that of s.c. insulin therapy. As of to date no serious side effects have been reported from these human trials. In summary, it appear that after several decades of research, for the first time a feasible alternative route for insulin administration is within reach.