Striga asiatica (L.) Kuntze, or witchweed, causes major yield reductions in principal cereal grain crops of Zimbabwe, including sorghum [Sorghum bicolor (L.) Moench]. Yield losses can be up to 100% in heavily infested fields. The development of high yielding cultivars with durable resistance to Striga (Striga spp) is the most feasible method for achieving sustainable control of this parasite in smallholder agriculture. Development of Striga resistant cultivars by conventional breeding is however slow and has been hampered by the lack of efficient and reliable screening techniques in breeding programs. Identification of molecular markers that are linked to witchweed resistance can expedite the development of resistant cultivars through adoption of appropriate marker assisted selection (MAS) strategies. We therefore sought to identify molecular markers that are linked to S. asiatica seed germination stimulant production, which is one the recognised mechanisms of resistance to witchweeds. A segregating F2 population derived from crosses between cultivars SV-1 (high germination stimulant producer, Striga-susceptible) and SAR 29 (low germination stimulant producer, Striga-resistant) was used for this purpose. Parental and F2 genotypes were screened for S. asistica seed germination stimulant production using the agar gel technique (AGT). The index of resistance was the maximum distance from the sorghum root at which S. asiatica seeds were found germinating. Genotypes with a germination distance of less than 1 cm were classified as resistant and those with a distance greater than or equal to 1 cm were classified as susceptible. DNA of AGT-classified genotypes were then assayed using random amplified polymorphic DNA (RAPD) and simple sequence repeat (SSR) markers. Linkage analysis was performed using the computer software MAPMAKER 3.0b, at a threshold LOD score of 3.0 and a 0.50 maximum recombination frequency. Segregation ratios of high to low F2 stimulant producers did not differ significantly (P £ 0.05) from the expected ratio of 3:1. It was therefore concluded that a single recessive gene controls low witchweed seed germination stimulant production in cultivar SAR 29. A total of 68 marker loci (10 SSR and 58 RAPD) were scored in the segregating F2 population. Segregation analysis of individual marker loci using the chi-square goodness-of-fit test revealed that 43 (62.32%) of the loci segregated normally while 26 (37.68%; 2 SSR and 24 RAPD) exhibited distorted segregation (P £ 0.05). Cosegregation analysis resulted in 39 markers (3 SSR and 36 RAPD) being assigned to nine linkage groups spanning a total distance of 817.1 cM. Thirty loci, including the locus for low stimulant production, were not assigned to any linkage group. The map was constructed from an average of 60 individuals per primer. The average, minimal and maximal distances between markers were 11.8 cM, 3.2 cM and 176.1 cM, respectively. More RAPD and SSR markers are currently being added to the molecular linkage map. This will improve coverage of the genome and facilitate identification of molecular markers for S. asiatica resistance in sorghum.