The phlebotomine sand fly Lutzomyia serrana (Damasceno & Arouck) was mass-reared under conditions of varying densities in an effort to improve colony production efficiency. To do this, the experimental carrying capacity of a standard rearing chamber was determined, i.e., the optimum population size in relation to density (individuals per unit of space). Rearing chambers of 100 cm³ were populated with 1–50 L. serrana engorged females and an equal number of males. Laboratory conditions were maintained at 23–26°C and 85–95% RH. The following parameters were recorded for each experimental chamber (three replicates): (1) female mortality without oviposition, (2) number of eggs oviposited and (3) number of adults emerging from the egg cohort. Female mortality began to increase substantially in the 26-female chamber, from 5.7% to 15% and finally reaching 60.2% in the 46–50 female chambers. In the chambers containing 1–20 females, egg number and realized adult progeny increased linearly to reach an asymptote. In the 20–50 female chambers, the number of eggs ranged from 420 to 699, and adult production from 306 to 432. The optimum carrying capacity for the 100-cm³ chambers was 22 ± 2 females. Beyond this number, auto-regulation was initiated, i.e., female mortality without oviposition increased as the number of females per chamber increased. Total number of eggs and adult production was similar in all chambers containing 20–50 females. In conclusion, for optimizing production of mass reared sand flies, determination of the carrying capacity is essential to optimize use of insectary resources, to avoid loss of valuable potentially ovipositing females, and to increase overall production efficiency.