All existing methods for analysing step-drawdown test data assume that the aquifer is homogeneous, whereas most real aquifers are heterogeneous. Failure to account for such heterogeneity can lead to erroneous mis-interpretation of step-drawdown test results, which may in turn result in expensive over- or under-design of long-term pumping plant. This paper reports a new method for analysing step-drawdown data which not only takes heterogeneity into account, but which also turns out to be simpler and more robust than existing methods even where the aquifer is relatively homogeneous. The new method is equally applicable to confined, leaky, and unconfined aquifers. The heart of this method is a simple data transformation procedure (which we term ‘homogenization of drawdown records’), based on an obvious logic, that unambiguously separates aquifer losses and well losses. As such, the new method directly calculates the well loss in the pumping well without recourse to the equations and coefficients upon which the existingmethods of analysis are based. Extending the same logic, it is apparent that well loss can also be quantified using data from constant-rate pumping tests. The new method not only yields more consistent and reliable results than existing methods, but also enhances conceptual understanding of drawdown processes and is sufficiently simple that it can be implemented rapidly by manual calculations, allowing analysis of test data in the field, even while pumping continues.