This study aimed to field tested the "Avoid the Needle" (AtN) intervention to reduce transitions from non-injecting to injecting drug use in two different epidemiological settings. Respondent driven sampling was used to recruit current non-injecting drug users (NIDUs) in Tallinn, Estonia in 2018-19 and in New York City (NYC) in 2019-20. Both persons who had never injected and persons who had previously injected but not in the last 6 months were eligible; a structured interview was administered, a blood sample collected, and the intervention administered by trained interventionists. We recruited 19 non-injectors from Tallinn and 140 from NYC. Participants in Tallinn were younger and had begun using drugs at earlier ages than participants in NYC. The primary drugs used in Tallinn were amphetamine, fentanyl, and opioid analgesics, while in NYC they were heroin, cocaine, speedball, and fentanyl. Six-month follow-up data were obtained from 95% of participants in Tallinn. The study was interrupted by COVID-19 lockdown in NYC, but follow-up data were obtained from 59% of participants. There were minimal transitions to injecting: 1/18 in Tallinn and 0/83 in NYC. There were significant declines in the frequencies of using readily injectable drugs (fentanyl, amphetamine, heroin, cocaine) from baseline to follow-up in both sites (Cochran-Armitage tests for trend, χ = 21.3, p < 0.001 for New York City; and χ = 3.9, p = 0.048 for Tallinn). Reducing transitions into injecting is a potentially very important method for reducing HIV transmission and other harms of drug use. Further investigation and implementation of AtN type interventions is warranted.