Previous studies have shown that the majority of long-lived cells harboring persistent HIV-1 proviral genomes originates from viruses circulating in the year prior to antiretroviral therapy (ART) initiation, but a smaller proportion originates from viruses circulating much earlier in untreated infection. These observations suggest that discrete biological factors influence the entry and persistence of viruses into the persistent proviral pool, and there may be periods earlier in untreated infection with increased seeding. Therefore, we examined the timing of formation of the long-lived pool of infected cells that persists during ART in seven women (after a median of 5.1 years of suppressive ART) by comparing the phylogenetic distance between unique 3' half genome on-ART proviral sequences and longitudinally sampled pre-ART viral RNA sequences, focusing on the period >1 year prior to ART initiation (i.e., the "early" proviral pool). We constructed models of continuous entry into the persistent proviral pool prior to ART initiation and analyzed the fit of our experimentally derived data to these models. We found that the pattern of persistent proviral pool formation in five of seven participants is incongruent with a model of continuous entry, implying that persistent proviral pool formation can occur episodically during untreated infection. Notably, increased entry into the persistent proviral pool was not universally observed during acute infection, and the timing of enhanced early entry differed across the participants.IMPORTANCECells harboring HIV-1 proviruses that persist on antiretroviral therapy (ART) constitute the main barrier to an HIV-1 cure. Recent work has elucidated that the majority of persisting proviruses harbor HIV-1 variants circulating near the time of ART initiation, whether the proviruses are intact or defective, though a portion forms earlier in untreated infection. We examined the formation of the "early-forming" persistent proviral pool and found that in 5/7 participants, persistent proviral pool formation was episodic, rather than continuous, suggesting that there are host/biological factors that periodically enhance the formation of the persistent proviral pool. Further characterization of these factors will aid in the development of methods to abrogate their effect, thereby reducing the size of the persistent proviral pool.