Background: Caffeine has been commonly used for prevention and treatment of apnea-related symptoms in premature infants. However, its side effects have not been thoroughly studied. We investigated whether caffeine affects gastric motility in very-preterm (VP) neonates. Methods: The study is a randomized crossover clinical trial. Twenty-two neonates with mean birth weight (BW) (standard deviation—SD) 1077 (229) g and mean gestational age (GA) (SD) 28.6 (2.1) weeks were recruited. Each neonate had its gastric emptying time checked twice with ultrasound assessment of changes in antral cross sectional area (ACSA). All neonates were sequentially allocated to the caffeine group (A) and the control group (B). Complications from the gastrointestinal tract were documented throughout the study. Results: Statistically significant difference was found with regards to the gastric emptying time [median, (range)] between caffeine and control group (p = 0.040). Additionally, in the neonates with BW 1000–1500 g and GA ≥ 28 weeks, the gastric emptying time (minutes) was significantly longer during caffeine treatment [44.5 (36–68.2)] and [40 (34.5–66.5)] respectively, as compared to the gastric emptying time during no caffeine treatment [27 (24.2–30)] (p = 0.002) and [27 (24.5–30)] (p = 0.001). The incidence of gastrointestinal (GI) complications was significantly greater in neonates receiving caffeine [6 (27.%)] as compared with those without caffeine treatment [1 (4.6%)] (p = 0.039). Conclusions: During caffeine treatment, a significantly delayed gastric emptying time was noted in all study neonates, especially in these with BW 1000–1500 g and those with GA ≥ 28 weeks. Further larger studies are necessary in order to confirm this interesting finding.