Light trap captures of larval, endangered Razorback Sucker Xyrauchen texanus are used to guide flow release timing from reservoirs and measure Razorback Sucker annual reproductive success in the Green River, Utah. However, little is known regarding efficacy of light traps to capture or retain larvae under variable and field-relevant environmental conditions. We conducted laboratory experiments to investigate the effects of light trap set time, fish release distance from trap, light presence, turbidity, fish density, and cover on capture and retention probabilities of five early life stages of Razorback Sucker. Mean capture probability of protolarvae prior to swim bladder development (7-9 mm TL) was 0.40 (range = 0.28-0.55) over various treatments but rose to 0.76 (0.73-0.80) after swim bladder development in later protolarvae (9-10 mm TL). Mesolarvae (11-17 mm TL), the most common life stage captured in field sampling, had a similar mean capture probability (0.86; 0.82-0.90) as later protolarvae. Capture probabilities of metalarval (mean = 0.42, range = 0.21-0.63; 15-24 mm TL) and juvenile (mean = 0.24, range = 0.20-0.28; 22-37 mm TL) life stages were lower. Larvae retention probabilities were generally > 0.75 and increased to 0.97 for juveniles. Longer set times positively influenced capture probabilities, though capture probabilities were similar in 1-, 3-, and 5m release distances and relatively high (>0.75) for most the commonly sampled life stages in standard conditions. Light presence in traps greatly increased capture and retention of larvae compared with unlit traps. Our results confirm that light traps may be useful to detect first presence of Razorback Sucker larvae in Green River backwaters each spring, which triggers Flaming Gorge Reservoir releases to inundate floodplain wetlands and for annual monitoring of reproductive success. We also provide recommendations for gear set up, deployment, and techniques to increase capture and retention probabilities of Razorback Sucker larvae.