Step-drawdown tests are usually carried out with succeeding steps of increased pumping yield, each step of equal duration of 1 hour. The background for this procedure is reviewed, and it is shown, that very often much shorter duration of steps is adequate.

The analysis of step-tests is usually based on the formulas of Jacob or Rorabaugh. The physical background for these formulas is reviewed, and the latter is shown to be unsatisfactory. However, the formula is used in cases where the data do not fit the Jacob formula, some typical deviations are shown. Reasons for these problems, trivial and non-trivial, are discussed, and some reasonably sound pshysical explanations for the various systematic deviations from the Jacob formula are given.