Stopped-flow time-scan experiments on both Triton X-100 (TX100) micelle and sodium dodecylsulfate (SDS) micelles, with the pyrene-containing triglyceride 1 as a probe, establish that there are two distinct solute exchange mechanisms with rates on the time scale of milliseconds to minutes. One process exhibits second order kinetics with a rate proportional to the concentration of empty micelles. For TX100 micelles, this process is rapid (k2 approximately 10(6) M(-1) s(-1) at 24.6 degrees C) and is characterized by an activation energy of 160 kJ mol(-1). From the fact that this rate is nearly independent of the structure of the probe we infer that the exchange involves micelle fusion to form a short-lived super-micelle, followed by fragmentation to form two normal (or 'proper') micelles. The rate of the first-order process decreases as the size of the probe increases (1-octylpyrene > 1-dodecylpyrene > 1). For SDS, both rates are very sensitive to the salt (NaCl) concentration. All indications point to this exchange process involving rate-limiting fragmentation of the micelle into two sub-micelles, these in turn grow back to normal micelles by addition of surfactant monomers or by collision with other sub-micelles. We explain the dependence of this rate on the nature of the probe by suggesting that only sub-micelles of a certain size are capable of carrying the probe with them as they separate from the original micelle.