A new production approach to make microbubbles (MBs) loaded with nanoparticles (NPs) (Protocol 1) was evaluated and compared with the more common procedure that was based on covalent linking of functionalized NPs (f-NPs) to MBs (Protocol 2). MBs were produced by sonicating bovine serum albumin (BSA) and dextrose solution. NPs consisted of Cy5-PLA conjugate. Protocol 1 involved incorporating the NPs into the BSA-dextrose solution before the sonication step. Protocol 2 involved mixing MBs and f-NPs, resulting from mixing the initial Cy5-PLA and PLA-PEG-COOH conjugates. For each protocol, unloaded MBs were produced as a reference. Two parameters were quantitatively analyzed using both analysis of variance (ANOVA) and t-test: diameter was estimated using a circle detection routine based on the Hough transform while the number density was estimated using a hemocytometer. Both parameters were evaluated for the NP-loaded MBs and unloaded MBs at 5 time points (2 hours to 5 days post MB fabrication). Protocols 1 and 2 resulted in significantly different loaded MBs: the more common approach of linking functionalized NPs to the MB surface (Protocol 2) was much more efficient than directly embedding NPs in the MB shell.