Subjective and objective semen assessments were performed on 18 male Mohor gazelles (Gazella dama mhorr). Sperm motility assessments combined with sperm plasma membrane and acrosomal integrity evaluations were undertaken as part of a captive breeding programme. The primary objective was to test methodology for short-term preservation of gazelle semen for artificial insemination (storage in N-[Tris(hydroxymethyl)methyl]-2-aminoethanesulphonic acid-Tris diluent (TEST) for up to 96 h at 17 degrees C). However, the secondary objective was to investigate phenotypic and genotypic influences on semen quality within this small population, which was established in 1971 with only 12 genetic founders. Sperm motility was measured by computer-assisted semen assessment and the data were analysed using a pattern analysis technique to detect and quantify naturally occurring sperm subpopulations within the semen samples. Four sperm subpopulations distinguishable by their motion characteristics were detected. The relative frequencies of two subpopulations (population 2: highly motile, non-linear; and population 4: poorly motile, non-linear) in fresh semen were correlated with the maximum voltage used during electroejaculation. The frequency of subpopulation 2 was negatively correlated with maximum voltage (r = -0.875, P < 0.0001) and the frequency of subpopulation 4 was positively correlated (r = 0.727, P < 0.005). The frequencies of all subpopulations varied significantly among the animals sampled (chi-squared = 2577.6, degrees of freedom = 54, P < 0.0001) and subpopulation 4 was also correlated with body weight (r = -0.59, P < 0.005). Semen stored at 17 degrees C retained motility, plasma membrane and acrosomal integrity for 48 h, but these measures decreased thereafter. The frequency of a sperm subpopulation showing uncoordinated but active motility increased significantly over the first 48 h and then decreased.