Automated Planning is a pivotal field of artificial intelligence, focusing on intelligent agents’ ability to generate action sequences leading from an initial state to a desired goal condition. A well-known technique to improve planning performance is based on macro-actions, which can reduce search depth by merging multiple primitive actions together, generating “shortcuts” in the search space. Macros have been studied extensively in classical planning, but rarely in more expressive formalisms. In this study, we investigate macro-actions in numeric planning, formalising the macro generation process and exploring a semi-automated methodology for selecting candidate primitive actions to be combined into macro-actions. Our extensive experimental analysis demonstrates the potential benefits of macros for numeric planning engines, providing useful insights into their effectiveness for efficient plan generation.