We show that spacing statistics can be obtained for all fracture densities by the same equation. Using the simple ‘shear lag’ theory, it is demonstrated that the results adequately fit experimental measurements. Theoretically, it is shown that two parameters are sufficient to characterize all spacing distributions. It is found that no real spacing saturation exists. Rather, a very low increase in joint densities occurs for a very high incremental stress increase. Moreover the transition to this kind of saturation is gradual and not abrupt. In this way the high density distributions with very small inter-joint distances, which pose a problem for saturation models, are directly explained. Correlation with creep experiments is also established. Comparison of our method with Weibull statistics is performed.