Hyperactivity, a form of sperm motility characterized by vigorous flagellar movements, has been proposed as essential for fertilization in mammals. The objective of the present study was to establish a method for inducing hyperactivity in vitro in boar spermatozoa and to define threshold values to differentiate between hyperactive and non-hyperactive spermatozoa by computer-assisted sperm analysis (CASA) as a prerequisite for analyzing the energy metabolism during hyperactivity. In TALP-HEPES medium, non-frozen boar spermatozoa were stimulated to hyperactivity by 50 mmol l to-head agglutination were low so that hyperactive spermatozoa could be analyzed for at least 40 min. The transition from normal to hyperactive movement was characterized by an increase in flagellar beat angle from 498 6 128 to 2008 6 368 (n 5 32) and a decrease in flagellar curvature ratio from 0.89 6 0.04 to 0.47 6 0.11 (n 5 32). For quantification of hyperactive boar sperm, kinematic parameters of hyperactive and non-hyperactive spermatozoa were measured by CASA and statistically evaluated (receiver operating characteristic (ROC) curve analysis). The threshold values of the following four parameters were well suited for differentiating between hyperactive and non-hyperactive boar spermatozoa (ROC curve analysis: >50% specificity at 100% sensitivity). Hyperactive boar spermatozoa showed mean lateral head displacement >3.5 mm, curvilinear velocity >97 mm s
21, linearity < 32% and wobble <71%. According to this multiparametric definition, induction of hyperactivity increased significantly (P < 0.0001) the fraction of hyperactive spermatozoa in semen samples from 5.1 6 4.3% (n 5 13) to 48.3 6 6.6% (n 5 7) in the absence and to 44.2 6 7.6% (n 5 10) in the presence of 25% seminal plasma, while the overall percentage of motile spermatozoa did not change significantly.