The paper presents a simplified version of the Arakawa-Schubert (AS) cumulus parameterization, called Relaxed Arakawa-Schubert (RAS). The key simplification in RAS is to relax the state toward equilibrium rather than requiring a balanced final state, as in the original AS implementation. This approach is more efficient and simpler to code, and avoids an ill-posed problem in AS due to the need to solve for a balanced state. RAS is evaluated using data from the GARP Atlantic Tropical Experiment (GATE) Phase III and is found to produce results equivalent to the standard AS implementation but with greater efficiency. The RAS parameterization simplifies the entrainment relation and relaxes the state toward equilibrium. It treats each cloud type separately, allowing a fraction of the mass flux to affect the sounding, and iteratively adjusts the cloud work function. The cloud work function is expressed in terms of grid-scale variables, and the parameterization is applied to vertically discrete models. The RAS approach uses a linearized entrainment relation and allows for a more efficient computation by considering the time scale of the adjustment. The paper compares RAS with the standard AS implementation, showing that RAS produces smaller temperature and moisture changes and requires less mass exchange for adjustment. RAS also avoids overadjustment of certain cloud types, which is a common issue in the standard implementation. The RAS parameterization is more computationally efficient, taking approximately 5 seconds per simulated day on a single CRAY-YMP processor, compared to 80 seconds for the standard implementation. This efficiency is attributed to the linearization of the entrainment relation and the simplification of the solution method. The semiprognostic evaluation of RAS using GATE Phase III data shows that it can accurately predict cumulus precipitation and atmospheric warming and drying. The results indicate that RAS is a viable alternative to the standard AS implementation, offering greater efficiency and simplicity while maintaining accuracy. The parameterization is particularly useful for climate and numerical weather prediction models that require the Arakawa-Schubert parameterization.The paper presents a simplified version of the Arakawa-Schubert (AS) cumulus parameterization, called Relaxed Arakawa-Schubert (RAS). The key simplification in RAS is to relax the state toward equilibrium rather than requiring a balanced final state, as in the original AS implementation. This approach is more efficient and simpler to code, and avoids an ill-posed problem in AS due to the need to solve for a balanced state. RAS is evaluated using data from the GARP Atlantic Tropical Experiment (GATE) Phase III and is found to produce results equivalent to the standard AS implementation but with greater efficiency. The RAS parameterization simplifies the entrainment relation and relaxes the state toward equilibrium. It treats each cloud type separately, allowing a fraction of the mass flux to affect the sounding, and iteratively adjusts the cloud work function. The cloud work function is expressed in terms of grid-scale variables, and the parameterization is applied to vertically discrete models. The RAS approach uses a linearized entrainment relation and allows for a more efficient computation by considering the time scale of the adjustment. The paper compares RAS with the standard AS implementation, showing that RAS produces smaller temperature and moisture changes and requires less mass exchange for adjustment. RAS also avoids overadjustment of certain cloud types, which is a common issue in the standard implementation. The RAS parameterization is more computationally efficient, taking approximately 5 seconds per simulated day on a single CRAY-YMP processor, compared to 80 seconds for the standard implementation. This efficiency is attributed to the linearization of the entrainment relation and the simplification of the solution method. The semiprognostic evaluation of RAS using GATE Phase III data shows that it can accurately predict cumulus precipitation and atmospheric warming and drying. The results indicate that RAS is a viable alternative to the standard AS implementation, offering greater efficiency and simplicity while maintaining accuracy. The parameterization is particularly useful for climate and numerical weather prediction models that require the Arakawa-Schubert parameterization.