Background. The optimal time to initiate hands-on training in endoscopic ultrasound fine needle aspiration (EUS-FNA) is unclear. We studied the feasibility of initiating EUS-FNA training concurrent with EUS training. Methods. Three supervised trainees were instructed on EUS-FNA technique and allowed hands-on exposure from the onset of training. The trainee and attending each performed passes in no particular order. During trainee FNA, the attending provided verbal instruction as needed but no hands-on assistance. A blinded cytopathologist assessed the adequacy (cellularity) and diagnostic yield of individual passes. Primary outcomes compared cellularity and diagnostic yield of attending versus fellow FNA passes. Results. We analyzed 305 FNA sites, including pancreas (51.2), mediastinal/upper abdominal lymph node (LN) (28.5) and others (20.3). The average proportion of fellow passes with AC was similar to attending FNApancreas: 70.3 versus 68.8; LN: 79.0 versus 81.7; others 65.5 versus 68.7; P > 0.05); these did not change significantly during the training period. Among cases with confirmed malignancy (n = 179), the sensitivity of EUS-FNA was 78.8 (68.4 fellow-only versus 69.6 attending only). There were no EUS-FNA complications. Conclusions. When initiated at the onset of EUS training, attending-supervised, trainee-directed FNA is safe and has comparable performance characteristics to attending FNA.