... Therefore caution should be taken when relying on ownerreported body condition scores in investigations of risk factors for feline obesity, and should not be used when assessing prevalence. Factors which increase the risk of feline overweight/obesity identified from previous studies include being male (Robertson, 1999;Lund et al., 2005;McGreevy et al., 2008;Colliard et al., 2009;Courcier et al., 2012), neutered (Sloth, 1992;Scarlett et al., 1994;Fettman et al., 1997;Robertson, 1999;Russell et al., 2000;Lund et al., 2005;Colliard et al., 2009;Courcier et al., 2010Courcier et al., , 2012), mixed breed or non-pedigree (Scarlett et al., 1994;Robertson, 1999;Lund et al., 2005;McGreevy et al., 2008), living in single-or two-cat households (Robertson, 1999), confinement indoors (Sloth, 1992;Scarlett et al., 1994;Robertson, 1999), living in a rural or semi-rural area (McGreevy et al., 2008), inactivity (Scarlett et al., 1994), middle age (Sloth, 1992;Scarlett et al., 1994;Harper, 1998;Russell et al., 2000;Kienzle and Bergler, 2006;McGreevy et al., 2008;Colliard et al., 2009;Courcier et al., 2012), increasing age (Cave et al., 2012) longer leg length (Allan et al., 2000;Cave et al., 2012), owner underestimation of their cat's body condition (Allan et al., 2000;Colliard et al., 2009;Cave et al., 2012), feeding ad libitum or free-choice (both terms used to describe food made continuously available) (Russell et al., 2000;Harper et al., 2001;Kienzle and Bergler, 2006), feeding in meals two or three times a day (Courcier et al., 2010), feeding treats two-three times a week (Russell et al., 2000), premium or therapeutic food (Lund et al., 2005), fresh meat, kitchen scraps or additional treats added to a cat's regular food, using extra food or a cat's favourite dish as a treat instead of play, more frequently providing food rewards when cat begs for food, watching a cat eating, and a close owner-cat relationship (Kienzle and Bergler, 2006). These previous studies have all been cross-sectional in design, resulting in uncertainty over the causal direction of associations observed between putative risk factors and obesity. ...