Slatted concrete floors are commonly used in sow housing for ease of manure handling and durability. The floor configuration and properties may markedly affect animal welfare and health. This research investigated the slatted concrete floors in sow barns in terms of their performance in manure drainage, surface friction, and sow gait. In-barn tests were conducted to evaluate a slatted floor configuration which had 105 mm wide slats and 19 mm slots and was compared with the most commonly used configuration of 125 mm wide slats and 25 mm slots. Digital images were taken to assess manure drainage through the slatted floors and ammonia concentrations were measured in the rooms as an air quality indicator. The surface roughness and dynamic coefficient of friction (DCOF) of the two concrete floors were measured as well. The results showed there was no significant difference in air quality, floor and animal cleanliness between the two floor configurations. The friction tests revealed that the length of time of floor usage by sows had a significant effect on the roughness and DCOF of concrete floors, with a sharp reduction in DCOF after the first week of use because manure stuck in the pores of the concrete surface reduced the interlocking between the asperities of the contact surfaces. A two-dimensional (2D) stick model connected by thirteen body joints was developed to simulate sow walking on the concrete floors. The video images of sows with reflective markers on the bodies walking along a corridor of concrete floor were analyzed and translated into a 2D coordinate dataset. The displacements of each joint at a time step in x- and y-directions were modelled by either a 2nd order polynomial function or a linear function. The predicted joint coordinates at each time step were then used to calculate a selected set of characteristic gait parameters. Based on the comparisons of the predicted gait parameters between the lame and non-lame sows, the criteria were established to detect sow lameness. The walking speed and forelimb stride length combination was defined as the best model indicator for lameness detection with an accuracy greater than 92%.