A process for producing hollow microcapsules as ultrasound contrast agents was optimized using a 2(3) factorial experimental design method with two replicates. Spray drying, a conveniently scalable encapsulation technique, was used to encapsulate a volatile core material, such as ammonium carbonate, using biodegradable 50-50 poly(D,L-lactide-co-glycolide). Various effects due to changes in processing variables and their interactions were studied using the factorial grid. The high- and low-incremented variables examined included the temperature difference between the inlet and outlet of the spray dryer (5 degrees and 15 degrees C), air atomization pressure (80 and 100 psi), and polymer concentration in solvent (0.005 and 0.025 g/mL). Responses analyzed for computing the main effects and interactions were microcapsule morphology, yield, mean size, and zeta potential. Experimental results showed that polymer concentration was most important for determining microcapsule morphology. The temperature difference for drying prominently affected mean size, and atomization pressure was the main effect for microcapsule yield. Interactions among variables were not present in this case. The best conditions for producing PLGA microcapsules was a temperature difference of 5 degrees C, an initial polymer concentration of 0.005 g/mL, and an atomization pressure of 80 psi. The microcapsule zeta potentials were unaffected by spray-drying conditions.