Long-term monogamy is most prevalent in birds but is also found in lizards. We combined a 31-year field study of the long-lived, monogamous Australian sleepy lizard, Tiliqua rugosa, with continuous behavioural observations through GPS data logging, in 1 yr, to investigate the duration of pair bonds, rates of partner change and whether either the reproductive performance hypothesis or the mate familiarity hypothesis could explain this remarkable long-term monogamy. The reproductive performance hypothesis predicts higher reproductive success in more experienced parents, whereas the mate familiarity hypothesis suggests that effects of partner familiarity select for partner retention and long-term monogamy. Rates of partner change were below 34% over a 5-yr period and most sleepy lizards formed long-term pair bonds: 31 partnerships lasted for more than 15 yr, 110 for more than 10 yr, and the recorded maximum was 27 yr (ongoing). In the year when we conducted detailed observations, familiar pairs mated significantly earlier than unfamiliar pairs. Previous pairing experience (total number of years paired with previous partners) had no significant effect. Early mating often equates to higher reproductive success, and we infer that is the case in sleepy lizards. Early mating of familiar pairs was not due to better body condition. We propose two suggestions about the proximate mechanisms that may allow familiar pair partners to mate earlier than unfamiliar partners. First, they may have improved coordination of their reproductive sexual cycles to reach receptivity earlier and thereby maximise fertilisation success. Second, they may forage more efficiently, benefiting from effective information transfer and/or cooperative predator detection. Those ideas need empirical testing in the future. Regardless of the mechanism, our observations of sleepy lizard pairing behaviour support the mate familiarity hypothesis, but not the reproductive performance hypothesis, as an explanation for its long-term monogamous mating system.