Speciation by sexual selection research has traditionally concentrated on mechanisms for divergence driven via female mate choice (intersexual selection). The pivotal role of competition between members of the same sex (intrasexual selection) has been largely overlooked. In this thesis, I describe a series of experimental studies investigating the role of intrasexual selection in sympatric divergence, using Lake Victoria haplochromine cichlid fish as a model system. Such experiments are needed, as only indirect or descriptive support exists for the hypothesis of male-male competition facilitating sympatric divergence. I also investigate color expression as a result of male-male competition, and ask if the intensity of sexual signaling provides information about the genetic quality of the bearer. Speciation theory suggests that ‘good genes’ mechanism of female mate choice are less likely to drive (sympatric) speciation than purely divergent Fisherian runaway selection (Lande 1981; Payne & Krakauer 1997; Turner & Burrows 1995; Van Doorn et al. 1998; 2004). Under Fisherian runaway selection, population bifurcation may proceed in arbitrary directions, whereas ‘good genes’ sexual selection is less arbritary, leading to the evolution of only those male traits that reliably indicate genetic quality. Identifying the mechanisms that cause sexual selection in haplochromines has important implications for our understanding of sympatric speciation. In chapter 2-7 I studied two recently diverged, ecologically and anatomically similar sympatric cichlid species pairs consisting of P. Pundamilia (Seehausen et al. 1998a) with blue and P. nyererei (Witte-Maas & Witte 1985) with red nuptial coloration. In chapter 3 I also use P. ‘pink anal’ (Seehausen 1996) as a representative of a Pundamilia species with blue nuptial coloration. Red and blue phenotypes are anatomically similar forms that behave like reproductively isolated species in some locations, and like hybridizing incipient species in other locations (Seehausen 1996; Seehausen et al. 1997). Blue phenotypes have a lake-wide distribution whereas red phenotypes have a patchy distribution and always co-occur with blue phenotypes. The blue form has the highest record of sympatric occurrences with other members of the Pundamilia complex (Seehausen 1996; Seehausen & Van Alphen 1999). It seems hence likely that blue represents the ancestral state and that blue populations have been invaded repeatedly and independently by red morphs during speciation (Seehausen 1997; Seehausen & Van Alphen 1999; Seehausen & Schluter 2004). The different populations of Pundamilia were taken as different speciation stages, beginning with an entirely blue population where red males had not been able to gain a foothold yet, to populations where red and blue are hybridizing incipient species, and finally to populations where speciation has been completed with reproductively isolated red and blue species. Own-type biases in aggression (chapter 2 and 3) and differences in aggression level (chapter 2) were measured in wild-caught Pundamilia males using a simulated intruder choice test. Territorial defenders were presented with red and blue stimulus fish, each enclosed in transparent tubes, mimicking intruding rival males, and recorded the aggressive response of the territorial defender. In chapter 2 light manipulations experiments revealed that aggression preferences are based on color differences, suggesting that color is not only important in mate choice, but also in intrasexual signaling. Following up to this chapter, in chapter 3 I tested in blue Pundamilia fish from six different populations whether aggression bias depends on speciation stage. Color-effects on dominance can be measured by staging dyadic combats, mimicking two males disputing over a vacant territory. In chapter 4, I tested the hypothesis that red coloration confers an advantage in direct combat, assisting red phenotypes to invade. To this end, combats were staged between red and blue males under both white and green light condition that effectively eliminates the color difference. How aggression biases come about in Pundamilia is unclear. Chapter 5 investigates the role of learning and genetics in shaping aggression biases in three populations of blue Pundamilia cichlid species. Males were given prolonged experience with red males or only blue males and subsequently subjected to a simulated intruder choice test.