Meiosis is too complex to have arisen at once full blown and a stepwise scheme is proposed for its evolution, where each step is believed to have provided an immediate selective advantage: (1) The first step in this tentative sequence is the development of a haploidization process by means of a rapid series of mitotic non-disjunctions, turned on under conditions where haploidy is favored. The non-disjunctions may have resulted from a conditional mutation which caused sister centromere cohesiveness in the past mitotic metaphase. (2) Next probably came the formation of rudimentary synaptonemal complex type structures, first at Holliday-type configurations and later extending from these along chromosome pairs. These structures between homologues, though costly to produce and maintain, may have directly served the disjunctive function by setting the stage for the production of haploidy in one division, under conditions where it was advantageous. (3) Then secondarily acquired functions of the synaptonemal complex or structures associated with it may have promoted greatly increased crossover frequency, in part at least by increasing the frequency of the isomerization-type reaction,. The resulting recombination of linked genes could have been advantageous under some conditions. (4) Finally, it is proposed that the capability was acquired for enhanced association of sister chromatids during the period between pachytene and anaphase I to give rise to chiasma-mediated disjunction, so that the relatively costly synaptonemal complex maintenance until anaphase I could be abandoned without losing disjunctive capability. It is implied that the modern synaptonemal complex is a structure which embodies a number of separately encoded proteins and that secondary structures and functions are associated with close homologue pairing. This scheme is based upon observable cytological and molecular characteristics of modern organisms.