To test the effect of stocking density on growth, survival rates and size distribution of juvenile swordtails, Xiphophorus helleri Heckel, under different environmental conditions, fish were stocked at rates of 1, 3 and 6 fish I−1. Growth was monitored over a 14-week period. Fish were weighed in six groups of 20 fish per tank. Size variation was determined at the end of the growth period by measuring the total length of each fish to the nearest mm. Fish were stocked in two different systems at three stocking densities (1, 3, and 6 fish I−1) in replicate. The systems were an indoor aquarium system and a tunnel system under semicommercial conditions. In each system, increasing stocking rates resulted in a significant reduction of growth rates. In the aquarium system, survival rate was not related to stocking density, whereas in the tunnel system, fish at the highest stocking density had a lower survival rate than those kept at the two lower stocking densities. The number of marketable fish above a set size limit of 40 mm total length was not higher in fish stocked at 6 fish I−1 than at 3 fish I−1 at the end of the study. Therefore, a stocking rate of 3 fish I−1 would be optimal. At the end of the study, fish were categorized into males, females and immature fish. Fish with a clearly visible gonopodium were classified as males, those with typical female characteristics as females, and small fish that were not yet differentiated as immature fish. The ratio of females to males was highly skewed in both systems, favouring a higher percentage of females. As stocking density increased, the female:male ratio was significantly reduced, which would result in a greater number of saleable pairs at increased stocking density. The implications of these results for the management of intensive production of swordtails are discussed.