As part of a map-based cloning strategy designed to isolate the root-knot nematode resistance gene Mi, tomato F2 populations were analyzed in order to identify recombination points close to this economically important gene. A total of 21,089 F2 progeny plants were screened using morphological markers. An additional 1887 F2 were screened using PCR-based flanking markers. Fine-structure mapping of recombinants with newly developed AFLP markers, and RFLP markers derived from physically mapped cosmid subclones, localized Mi to a genomic region of about 550 kb. The low frequency of recombinants indicated that recombination was generally suppressed in these crosses and that crossovers were restricted to particular regions. To circumvent this problem, a population of Lycopersicon peruvianum, the species from which Mi was originally introgressed, that was segregating for resistance was developed. Screening of this population with PCR, RFLP and AFLP markers identified several plants with crossovers near Mi. Recombination frequency was approximately eight-fold higher in the Mi region of the L. peruvianum cross. However, even within the wild species cross, recombination sites were not uniformly distributed in the region. By combining data from the L. esculentum and L. peruvianum recombinant analyses, it was possible to localize Mi to a region of the genome spanning less than 65 kb.