Although rapid on-site cytological evaluation (ROSE) for endoscopic ultrasound (EUS)-guided tissue acquisition (EUS-TA) may increase diagnostic yield, it is not widely available. Macroscopic on-site evaluation (MOSE) is an alternative modality although it is not standardized for EUS-guided fine-needle biopsy (FNB). We evaluated diagnostic performance of MOSE compared with conventional technique of EUS-TA using core biopsy needle. Consecutive patients undergoing EUS-FNA for solid lesions were randomized to MOSE or conventional arms. The primary and secondary outcome measures were diagnostic accuracy, diagnostic yield, sensitivity, specificity, positive and negative predictive values, and the number of passes, respectively. The optimum parameters for macroscopic visible core (MVC, i.e., length, number) by MOSE to achieve accurate diagnosis were evaluated. Ninety-six patients (48 conventional and 48 MOSE) were enrolled. Mean lesion size was larger in MOSE arm (32.67±7.22 vs 29.31±6.98mm, P=0.023). Diagnostic accuracy (95.8% vs 91.6%), diagnostic yield (97.9% vs 95.8%), procedure duration, and adverse events of the two methods were similar. Median number of passes with MOSE was less (2 vs 3 P=0.000). Area under the receiver operating characteristic curve showed that with MOSE, obtaining a total MVC length of 11.5mm had 93.3% sensitivity, and 2.5 MVC cores (each 4mm) had 86.7% sensitivity for malignancy diagnosis. EUS-FNB with MOSE, a simple reliable technique, can achieve a high and comparable diagnostic accuracy with lesser number of passes. Obtaining longer length and greater number of MVC increase the sensitivity to diagnose malignancy with MOSE.