Inducing protein aggregation in vitro under various formulation and stress conditions may lead to an increased understanding of the different association routes a protein can undergo. However, a range of factors can affect the aggregation process, often leading to heterogenous samples and experimental irreproducibility between labs. Here, we present detailed methods to reproducibly form homogenous samples of superstructures: amyloid-like fibrils, spherulites, and particulates from human insulin. We discuss pitfalls and good practice in the lab, with the aim of creating awareness on the potential sources of artefacts for protein stability and aggregation studies.