Apologies for cross posting Dear Madam/Sir: Refreshing my knowlegments on randomization process, i found the following: 1.- Ulrich Abel and Armin Koch, in an interesting paper entitled " The role of randomization in clinical studies: myths and beliefs" J Clin Epidemiol 1999;52:487-497, wrote " randomization is also not necessary for causal inference and it is also not necessary for hypothesis testing". In contrast, in the book entitled "Fundamentals of Clinicals Trials" of Friedman, Furberg and DeMets, page 61, chapter 5, appeared "...is the basis for statistical inference" and "...guarantees that statistical test will have valid significance". My ignorance is this field is very important, but my common sense, the less common of the sense, tell me that if the P depend of the randomization, then in any randomised controlled trial P will be <0.05, but it is not true. Who may to put light in my ignorance? Thank so much, warm regards from Valencia, Venezuela. Arturo Martì-Carvajal Universidad de Carabobo Venezuela [log in to unmask] fax:(58-41)-712883 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%