This work is aimed at the study of different methods for uncertainty evaluation in measurement, both theoretical and applied. In the metrological field, the internationally accepted approach for uncertainty evaluation implies use of the GUM (Guide to the Expression of Uncertainty in Measurement). There are however situations in which it may not perform satisfactorily, e.g. when the measurement function is non-linear. The JCGM (Joint Committee for Guides in Metrology) has recently described in Supplement 1 to the GUM an alternative method for uncertainty evaluation, that is propagation of distributions using a Monte Carlo method. This method requires however an appropriate knowledge of statistical distribution of input quantities, a hard to meet requirement. It is possible to remedy resorting to the bootstrap method (Efron 1979), which does not require prior assumptions. This method has been employed as an alternative solution for uncertainty assessment of CMM (Coordinate Measuring Machines) and subjective evaluations. Uncertainty contributions may be influenced by the presence of outliers; some outlier detection methods have thus been considered.