Measurements of aerosol number distributions down to one molecule have provided information that we’ve used to develop a new approach for modeling atmospheric nucleation rates. Measurements were carried out with the Cluster Chemical Ionization Mass Spectrometer (Cluster CIMS), the scanning mobility spectrometer using a diethylene glycol condensation particle counter as detector (DEG SMPS), and an ambient pressure proton transfer mass spectrometer for ammonia and amines (AmPMS). The model explains nucleation as a result of cluster evolution due to a sequence of acid-base reactions. We conclude that the smallest stable cluster contains four sulfuric acid molecules. The model leads to a simple analytic expression for nucleation rates that is reasonably consistent (i.e., ± 10x) with atmospheric observations. The model predicts that nucleation rates are equal to a prefactor, P<1, times the sulfuric acid vapor collision rate, (i.e., J=P⋅0.5⋅k11 ∗[H2SO4]2).