Given a finite or infinite word v, we consider the set M(v) of minimal forbidden factors of v. We show that the set M(v) is of fundamental importance in determining the structure of the word v. In the case of a finite word w we consider two parameters that are related to the size of M(w): the first counts the minimal forbidden factors of w and the second gives the length of the longest minimal forbidden factor of w. We derive sharp upper and lower bounds for both parameters. We prove also that the second parameter is related to the minimal period of the word w. We are further interested to the algorithmic point of view. Indeed, we design linear time algorithm for the following two problems: (i) given w, construct the set M(w) and, conversely, (ii) given M(w), reconstruct the word w. In the case of an infinite word x, we consider the following two functions: g_x that counts, for each n, the allowed factors of x of length n and f_x that counts, for each n, the minimal forbidden factors of x of length n. We address the following general problem: what information about the structure of x can be derived from the pair (g_x,f_x)? We prove that these two functions characterize, up to the automorphism exchanging the two letters, the language of factors of each single infinite Sturmian word.