Prototyping and optimizing of hydrocyclones for fine particle separation is as of now associated with high cost and long lead times due to reliance on injection molding, together with complex multivariable design dependencies. This study presents and validates an iterative design method aimed at improving separation performance while significantly reducing development time and cost. The method is grounded in a literature study and experimental validation using rapid prototyping. The work is focused on a hydrocyclone with a 10 mm cone diameter operating with a mixture of starch and water with majority of the starch particle sizes between 2.5 - 25 µm. Performance is evaluated based on two primary criteria, washing efficiency defined as minimizing the starch loss in the overflow and separation efficiency defined as maximizing starch concentration in the underflow. Affected factors as energy consumption and flow rate are excluded. A key contribution of this study is the validation of stereolithography (SLA) 3D-printing as a feasible prototyping method. Experimental results by comparing an SLA printed hydrocyclone against the injection molded benchmark hydrocyclone shows that the printed hydrocyclone even at high layer heights show comparable results, reducing the prototyping cost from approximately 43 000 SEK to around 20 SEK per prototype in combination with enabling short iteration cycles. The proposed method follows an adapted version of the spiral development process consisting of three stages, initialization, development, and finishing. Within this structure, key geometric parameters, cone profile, inlet design and vortex finder diameter are systematically explored and evaluated through experimental testing. The results demonstrate a successful implementation of rapid prototyping and an iterative method to  build up on in future development and an approach that can be applied in other complex geometry dependent systems.