In Physarum polycephslum, uninucleate haploid amoebae develop into macroscopic multinucleate plasmodia. Wild-type, sexual development is triggered when two amoebae carrying different alleles of mafA fuse to form a zygote which develops into a diploid plasmodium. Mutations in the mat4 genetic region give rise to apogamic strains in which a single haploid amoeba can develop into a haploid plasmodium. An essential stage in both sexual and apogamic plasmodium formation is an extended cell cycle in uninucleate cells, which ends with the formation of a binucleate cell by mitosis without cytokinesis. Using a 'brute force' screening method, we have isolated mutants blocked in apogamic plasmodium development. Genetic analysis showed that the mutations we have identified were unlinked to matA, unlike mutations previously identified following an enrichment step. Most of the loci revealed by our screen were represented by only one allele, indicating that further screening should lead to the identification of additional genes required for plasmodium development. Phenotypic analysis showed that different mutants were blocked at different stages of plasmodium formation. Some of the mutations blocking apogamic development at an early stage, close to the start of the long cell cycle, failed to block sexual development in zygotes homozygous for the mutation. Since the two modes of plasmodium formation differ only in the initiation of development, these mutations presumably interfere with the initiation process. In the remaining mutants, in which both sexual and apogamic development were blocked, development first became abnormal towards the end of the long cell cycle. This suggested that the wildtype gene products were required by this time and was consistent with previous evidence that many changes in cellular organization and gene expression occur during the long cell cycle. Each of these mutants showed a different terminal phenotype and some aspects of plasmodium development occurred normally although others were blocked, suggesting that development involves multiple pathways rather than a dependent sequence of events. Phenotypic analysis of double mutants supported this conclusion and also revealed epistatic interactions, presumably due to blocks in the same pathway. In several of the mutants, terminally differentiated cells died by an apoptosis-like mechanism; since this was never observed in vegetative cells, it was presumably triggered by the failure of development. Phenotypic analyses of additional mutants will extend our understanding of the pathways involved in plasmodium development.