Gene duplication is a process that produces extra copies of genes within the genome. Gene families are groups of similar genes, which evolutionist biologists believe to be products of gene duplication. In this paper, I argue that gene duplication is not the source of all of the modern gene families for the following reasons. (1) Most of the documented gene duplications are detrimental, and when beneficial, they cannot lead to new molecular functions. (2) Duplicated genes are usually silenced epigenetically, followed by degenerative mutations, ending up in non-functionalization. (3) Members of gene families are often components of irreducibly complex systems. (4) Regulation hierarchies, which have no counterparts in lower organisms, are required for proper expression of gene families. I propose the following criteria to distinguish between DNA sequences that were duplicated in history and paralogous genes that were created individually: First, copy-number polymorphisms among individuals of the same species obviously demonstrate recent duplications. Second, components of irreducibly complex systems are not likely products of gene duplications. These include genes with complex regulation hierarchies. Additional criteria are that the degree of sequence homology is a poor indictor to determine whether genes are duplicated, and the duplication by transposition is possible but is normally suppressed.