The present study aimed to evaluate the effect of the ultrasound technology (US) on microbial inactivation (MI) of a prebiotic whey soursop-flavored beverage compared with conventional method of rapid pasteurization (RP). The whey beverage was manufactured with whole pasteurized and homogenized milk (30 g 100 g⁻¹), whey powder (6 g 100 g⁻¹), soursop pulp (15 g 100 g⁻¹), sugar (8 g 100 g⁻¹) and inulin (6 g 100 g⁻¹). The US processing was realized in 25-cm³ samples, for 3 min at 3 different powers (200, 400 and 600 W), being one treatment submitted to RP (72–75°C for 15–20 s) and one was unprocessed (US 0 W), which was used as a control to calculate the log reductions. The whey beverages were submitted to microbial group counts (total and thermotolerant coliforms, aerobic mesophilic bacteria and yeasts counts) and some physico-chemical parameters (pH and zeta potential). Total and thermotolerant coliforms values were <3 MPN/mL in all samples, suggesting good hygienic quality during all processing while for aerobic mesophilic bacteria counts, the samples submitted to RP, US 600 W and US 400 W presented log reductions of 2, 2 and 1 respectively (P > 0.05), suggesting that higher US powers influence directly in MI. Finally, the yeasts counts presented absence of significant difference between US and RP (P > 0.05), despite the slightly higher reduction in US 600 W and 400 W treatments. pH values remained at 5.4 ± 0.01 regardless of the treatments (P > 0.05), while zeta-potential values of US samples ranged from −18 to −19 mV (US 200 and 600 W, respectively, P < 0.05), suggesting that the intensity of US powers changed the negative charges of the whey beverage particle surfaces; however no difference was observed when compared with US 0 W and RP (−19 mV for both processing, P > 0.05). Overall, US technology presented effective impact on microbial inactivation in a prebiotic soursop-flavored whey beverage.