This paper explores the conditions under which regression quantiles can be estimated by estimating conditional means. The authors propose an estimator that combines estimates of the location and scale functions, both identified by conditional expectations of appropriately defined variables. This approach allows the use of methods valid for conditional mean estimation while providing information on how regressors affect the entire conditional distribution. The method is particularly useful in panel data models with individual effects and models with endogenous explanatory variables. The paper presents the estimator, establishes regularity conditions for valid inference, performs a simulation study, and discusses two empirical applications. The authors argue that their approach complements existing quantile regression methods, making it easier to implement and applicable to a wider range of models, especially those with large datasets and multiple endogenous variables. The paper also highlights the importance of testing the assumption that covariates only affect the location and scale functions to ensure the validity of the approach in specific applications.This paper explores the conditions under which regression quantiles can be estimated by estimating conditional means. The authors propose an estimator that combines estimates of the location and scale functions, both identified by conditional expectations of appropriately defined variables. This approach allows the use of methods valid for conditional mean estimation while providing information on how regressors affect the entire conditional distribution. The method is particularly useful in panel data models with individual effects and models with endogenous explanatory variables. The paper presents the estimator, establishes regularity conditions for valid inference, performs a simulation study, and discusses two empirical applications. The authors argue that their approach complements existing quantile regression methods, making it easier to implement and applicable to a wider range of models, especially those with large datasets and multiple endogenous variables. The paper also highlights the importance of testing the assumption that covariates only affect the location and scale functions to ensure the validity of the approach in specific applications.