A simulation study was performed to investigate methods for mapping single-dose (simplex) and double-dose (duplex) markers, and for identification of homologous chromosomes in an autotetraploid species, and to see how the map accuracy depends on the population size. An initial population of 1000 individuals was simulated, with 30 simplex and 10 duplex markers, and recombination fractions and lod scores were calculated between all pairs of markers. These were used to test the feasibility of mapping the simplex and duplex markers simultaneously. Smaller populations, from 500 to 75 individuals, were then simulated, and the estimates of the pairwise recombination fractions and the derived maps were compared with the true map. It was found that the accuracy of the estimates depended strongly on the type of markers involved, with simplex–simplex coupling pairs being most reliable and simplex–simplex repulsion pairs and duplex–duplex pairs in any configuration but coupling being least reliable. Maps can be assembled using recombination fractions and lod scores from pairs of simplex–simplex markers (coupling and repulsion), duplex–simplex (coupling and repulsion) and duplex–duplex (coupling). The agreement between the map order and the true order was good, although the map distance was generally underestimated at small sample sizes.