Alternative techniques are considered to measure the validity of probabilistic precipitation forecasts from a hydrologic perspective. This includes techniques to measure informativeness or skill as well as techniques to measure reliability or calibration. Measures of informativeness or skill include measures based on a correlation parameter, ensemble mean rms error and the Brier skill score. Measures of reliability include uniformness of exceedance probabilities of observed precipitation events and comparisons of observed and forecast probability of precipitation. Measures of informativeness and skill depend on scale (averaging area and duration of forecast period), forecast lead time, location and season. Examples are presented by comparing archived global ensemble precipitation forecasts with observed precipitation. Statistics are presented before and after compensating for systematic model biases.