OBJECTIVES Medication non-adherence is often classified by the timing of non-adherence. Primary non-adherence is the failure to fill a newly-prescribed medication (Rx). Rx non-persistence is the failure to continue therapy after the initial fill. A recent classification − early non-persistence (ENP) − is the failure to obtain the first prescribed refill of a new Rx (single dispensation only). In this meta-analytic review, we compare the rates of ENP across studies by four moderator dimensions: (1) chronic disease class; (2) symptomatic vs. asymptomatic chronic condition; (3) length of the baseline treatment-free interval; and (4) whether ENP estimates were adjusted for Rx switches. METHODS Fifty-eight studies that contained data on ENP were identified using a PubMed literature search and searches of each article’s reference list. ENP rates were recorded for 91 distinct samples. ENP was defined as the failure to obtain the first prescribed refill within 30 days of the first-refill date. ENP rates were weighted by sample size and combined to provide pooled fixed effect size estimates for the moderator dimensions.  RESULTS Across all samples, the overall weighted ENP rate was 23.6%. Observed difference between ENP rates by disease class were largely explained by whether the treatment focused on symptom control or not: symptomatic ENP rate=39.5% vs. asymptomatic ENP rate=17.7%. ENP rates were affected by two aspects of methodology: (1) length of baseline treatment-free interval (shorter, favoring treatment-experienced=17.1% vs. longer, favoring treatment-naïve=27.2%); and (2) Rx switches accounted for in the ENP estimates (accounting for switch=18.8% vs. not accounting for switch=26.0%). CONCLUSIONS ENP is as high, and can be higher than primary non-adherence. Most extant studies simply document the rate of ENP. Given that ~ 24% of patients are “one-and-done,” it is imperative to: (1) understand the drivers of ENP and (2) develop patient-centered interventions to stem the epidemic of ENP.