We performed a field experiment to investigate the effect of carapace width, major cheliped length and burrow ownership on the righting success of male fiddler crabs (Uca annulipes). We removed males from their burrows and released them back into the colony (n= 82). Released males tended to initiate encounters with burrow owners slightly smaller than themselves. Several general predictions of Sequential Assessment Game models of contest behaviour were supported: (1) residents won more encounters; (2) intruders were more likely to win when larger than residents. When body size (carapace width) was controlled for, intruders with relatively large claws for their body size were more likely to win contests; (3) the duration of encounters was related to the size difference between males; (4) encounters won by the larger male were of shorter duration than those won by the smaller male; (5) encounters won by the resident tended to be of shorter duration than those won by intruders (P= 0.07); (6) on average, encounter duration was longer when the intruder was larger than the resident. However, the encounters we documented began with seemingly costly behaviour such as pushing and the inter-locking of claws and did not unambiguously escalate from initial low cost behaviours. Sequential assessment of relative fighting ability may therefore not have been occurring. Prior visual assessment of opponents' fighting ability, followed by ‘all-out fights’ during physical encounters may also provide a plausible explanation for our results.