Practical techniques for assessing semen quality in order to predict male fertility are still needed. The principal objective of this experiment was to evaluate routine laboratory evaluation and in vitro fertilization (IVF) techniques as predictors of relative boar fertility using a low-dose AI protocol. Nine boars were evaluated during a 6.5+/-1 mo period, beginning at 29-32 wk of age. Ejaculates were evaluated for motility, morphology and concentration, diluted to 1.5 billion sperm in 50 mL extender, and used to breed 50+/-5 gilts over the same period. On nine occasions, a specific aliquot of the ejaculate's first sperm-rich fraction was evaluated using IVF procedures. Boars differed (P<0.001) consistently for pregnancy rate (from 73 to 98%), farrowing rate (71-98%) and total born (8.8-12.0). Routine semen evaluation and IVF parameters that presented significant differences between boars, but no differences in time and no boar by time interaction, were used to correlate in vivo fertility. A multiple regression model based on routine semen evaluation parameters accounted for up to 27 and 22% of the variation of fertility index and total piglets born, respectively, whereas male pronuclear formation rate was the IVF variable that accounted for 17 and 12% of the variation in farrowing rate and fertility index, respectively. Collectively, we inferred that the use of low sperm numbers for AI, determination of pregnancy rate at Day 30, motility of extended semen after 7 and 10d, and specific IVF parameters may be useful for identifying relatively infertile boars that are not currently excluded from use in existing commercial boar studs.