Losses of pigs (dead and nonambulatory) during transport are of great concern from animal welfare and economic perspectives. Based on several field studies, the incidence of transport losses in market-weight pigs is approximately 1% (Ellis et al., 2003, 2004). Survey evidence suggests that overcrowding pigs during transport is associated with greater mortality rates (Robertson, 1987; Guardia et al., 1996; Riches et al., 1996). Currently, the National Institute for Animal Agriculture (2004) recommends floor space allowances of 0.40 to 0.45 m2/pig for pigs weighing between 114 to 136 kg (equivalent to approximately 0.35 and 0.33 m2/100 kg of BW, respectively). The objective of this study was to investigate effects of 2 floor spaces (0.39 and 0.48 m2/pig) during transport, which represent the range currently being used in commercial practice in the United States, on the incidence of dead and nonambulatory pigs and to evaluate relationships between transport conditions and losses. Data on 74 trailer loads of finishing pigs (mean BW = 129.0, SEM = 0.63 kg) from wean-to-finish buildings on 2 farms within 1 production system were collected to investigate the effect of amount of floor space on the trailer (0.39 or 0.48 m2/pig) during transport on the incidence of losses (dead and nonambulatory pigs) at the packing plant and to study the relationships between transport conditions and losses. Pigs were loaded using standard commercial procedures for pig handling and transportation. Two designs of flat-deck trailers with 2 decks were used. Floor space treatments were compared in 2 similarly sized compartments on each deck of each trailer type. Varying the number of pigs in each compartment created differences in floor space. The incidence of nonambulatory pigs at the farm during loading and at the plant after unloading, average load weight, load number within each day, event times, and temperature and relative humidity in the trailer from loading to unloading were recorded. Of the 12,511 pigs transported, 0.26% were nonambulatory at the farm, 0.23% were dead on arrival, and 0.85% were nonambulatory at the plant. Increasing transport floor space from 0.39 to 0.48 m2/pig reduced the percentage of total nonambulatory pigs (0.62 vs. 0.27 ± 0.13%, respectively; P < 0.05), nonambulatory, noninjured pigs (0.52 vs. 0.15 ± 0.11%, respectively; P < 0.01), and total losses (dead and nonambulatory pigs) at the plant (0.88 vs. 0.36 ± 0.16%, respectively; P < 0.05) and tended to reduce dead pigs (0.27 vs. 0.08 ± 0.08%, respectively; P = 0.06). However, transport floor space did not affect the percentage of nonambulatory, injured pigs at the plant. Nonambulatory pigs at the farm were positively correlated with relative humidity during loading and load number within the day (r = 0.46 and 0.25, respectively; P < 0.05). The percentage of total losses at the plant was positively correlated to waiting time at the plant, unloading time, and total time from loading to unloading (r = 0.24, 0.51, and 0.36, respectively; P < 0.05). Average temperature during loading, waiting at the farm, transport, waiting at the plant, unloading, and average pig weight on the trailer were not correlated to losses. These results suggest that floor space per pig on the trailer and transport conditions can affect transport losses.