Two explanations exist for the short-lived radionuclides (T 1/2 5 Myr) present in the solar system when the calcium-aluminum-rich inclusions (CAIs) first formed. They originated either from the ejecta of a supernova or by the in situ irradiation of nebular dust by energetic particles. With a half-life of only 53 days, 7 Be is then the key discriminant, since it can be made only by irradiation. Using the same irradiation model developed earlier by our group, we calculate the yield of 7 Be. Within model uncertainties associated mainly with nuclear cross sections, we obtain agreement with the experimental value. Moreover, if 7 Be and 10 Be have the same origin, the irradiation time must be short (a few to tens of years), and the proton flux must be of order F $ 2 ; 10 10 cm À2 s À1 . The X-wind model provides a natural astrophysical setting that gives the requisite conditions. In the same irradiation environment, 26 Al, 36 Cl, and 53 Mn are also generated at the measured levels within model uncertainties, provided that irradiation occurs under conditions reminiscent of solar impulsive events (steep energy spectra and high 3 He abundance). The decoupling of the 26 Al and 10 Be observed in some rare CAIs receives a quantitative explanation when rare gradual events (shallow energy spectra and low 3 He abundance) are considered. The yields of 41 Ca are compatible with an initial solar system value inferred from the measured initial 41 Ca / 40 Ca ratio and an estimate of the thermal metamorphism time (from Young et al.), alleviating the need for two-layer proto-CAIs. Finally, we show that the presence of supernova-produced 60 Fe in the solar accretion disk does not necessarily mean that other short-lived radionuclides have a stellar origin.