Long-tailed Manakins Chiroxiphia linearis are frugivorous birds with a lek mating system and male-male cooperation in courtship display. I studied male-male networks and correlates of male mating success in a color-banded population in Monteverde, Costa Rica, from 1981 to 1986. Males were organized in teams at scattered perch-zones (75 to 300 m apart) that were usually in aural but not visual contact. Each team consisted of 3 to 15 males (x=7.1±3.4), in an apparent linear dominance hierarchy, with an alpha and beta male who did most of the courtship display. In a study population with 50 to 60 active males per season, only 6 to 8 males were alphas. Only betas inherited alpha status (n=3). Males appear to be 8 or more years of age before attaining beta status. Alpha tenure can last 4 years. Alpha males were rarely or never seen in zones other than their 'home' zone. Lower-ranking males maintained simultaneous affiliations with males at as many as 6 different zones. Each zone, therefore, was a sort of hub at which males with different affiliations around the rim came into contact. Each of the 6 major perch-zones shared at least one constituent with each of the other zones. The mean number of males shared by zones was 3.9 ± 2.7 (range=1 to 9). Marked changes occurred in male traits with increasing age and status. These included (1) Significant declines in weight throughout the lifespan, (2) a 4-year delay in plumage maturation with well-defined stages, (3) reduction in the number of zones with which males maintained affiliations, and (4) increasing probability of copulatory success (restricted to a small subset of the oldest males, ≥ 10 years of age). Variance in copulatory success was the highest yet described for birds. Of 85 males monitored between 1983 and 1986, copulations (n=121) were distributed among only 8 males. Four of these males accounted for over 90% of the copulations, with 63% accruing to one male. The beta male of this alpha copulated twice in the absence of his partner; all the other copulaters were alphas. I examined correlates of male mating success. Female visitation correlated with the number of unison 'toledo' calls given. If a female visited, copulatory success correlated both with a residual effect of the 'toledo' output and with the duration of the 'butterfly' component of the dual-male dance performance. My correlational results suggest that females do choose, on the basis of performance cues, among the small subset of males that are well-established alpha and beta partners. Development of alliances, as much as male combat, may determine attainment of high-performance partner status. Thus, sequential male-male interactions and female choice appear to produce nested subsets of successful males leading to an extreme in variance in male mating success. Males unsuccessful in male-male interactions are not 'eligible' for female choice. By requiring partnered display, females may be implicitly narrowing the subset of potentially successful males. In other lek systems the union, rather than the intersection, of the subsets produced by intra- and intersexual selection may include successful males. In that case, intrasexual selection via disruption of copulations may enlarge the pool of potentially successful males under intersexual selection and produce lower variances in male mating success. Students of sexual selection may need to consider the extent to which intra- and intersexual selection interact as union or intersecting sets to produce variance in male mating success.