Much effort and progress have been made in understanding the nucleation and crystallization of globular proteins, and many techniques have been developed to crystallize proteins in the past decades. The advantages of the use of combined precipitants in protein crystallization have been much appreciated. Unfortunately, there is still no theory or empirical guide on how to combine so many precipitants and how to use combined precipitants, although many proteins have been crystallized successfully using combined precipitants. This report gives a proposal about how to use conventional precipitants to obtain protein crystals, based on a novel idea of hard and soft precipitant combinations.