In this paper, we provide three different ways to partition the polytope of doubly substochastic matrices into subpolytopes via the prescribed row and column sums, the sum of all elements and the sub-defect respectively. Then we characterize the extreme points of each type of convex subpolytopes. The relations of the extreme points of the subpolytopes in the three partitions are also given.Denote the set of all n × n doubly stochastic matrices by Ω n , and the set of all n × n doubly substochastic matrices by ω n . The set Ω n is a convex polytope, and has been intensively studied by many mathematicians [11,12,13,15,16,17,18,23,24,25,33,34,32]. Specially, the extreme points of Ω n are exactly the permutation matrices due to Birkhoff [4] and von Neumann [31], which can be stated as follows.Theorem 1.3. [4, 31] An n × n matrix A is a doubly stochastic matrix if and only if there are finite permutation matrices P 1 , P 2 , · · · , P N and positive numbers α 1 , · · · , α N such that α 1 + · · · + α N = 1 and A = α 1 P 1 + · · · + α N P N .The set ω n is also a convex polytope and its extreme points are partial permutation matrices [21], i.e., matrices with at most one element in each row and each column equal to one and other elements zero. Since Ω n ⊆ ω n , one may wonder if the classical 2010 Mathematics Subject Classification. 15B51; 52B05; 05A18.