Cell spheroids are promising three-dimensional (3D) models that have gained wide applications in many biological fields. This protocol presents a method for manufacturing high-quality and high-throughput cell spheroids using a 3D acoustic assembly device through maneuverable procedures. The acoustic assembly device consists of three lead zirconate titanate (PZT) transducers, each arranged in the X/Y/Z plane of a square polymethyl methacrylate (PMMA) chamber. This configuration enables the generation of a 3D dot-array pattern of levitated acoustic nodes (LANs) when three signals are applied. As a result, cells in the gelatin methacryloyl (GelMA) solution can be driven to the LANs, forming uniform cell aggregates in three dimensions. The GelMA solution is then UV-photocured and crosslinked to serve as a scaffold that supports the growth of cell aggregates. Finally, masses of matured spheroids are obtained and retrieved by subsequently dissolving the GelMA scaffolds under mild conditions. The proposed new 3D acoustic cell assembly device will enable the scale-up fabrication of cell spheroids, and even organoids, offering great potential technology in the biological field.