Trammel netting is generally the accepted method of monitoring razorback sucker in reservoirs, but this method is ineffective for monitoring this fish in rivers. Trammel nets set in the current become fouled with debris, and nets set in backwaters capture high numbers of nontarget species. Nontargeted fish composed 97 percent of fish captured in previous studies (1999-2005). In 2005, discovery of a large spawning aggregation of razorback sucker in midchannel near Needles, Calif., prompted the development of more effective methods to monitor this and possibly other riverine fish populations. This study examined the effectiveness of four methods of monitoring razorback sucker in a riverine environment. Hoop netting, electrofishing, boat surveys, and aerial photography were evaluated in terms of data accuracy, costs, stress on targeted fish, and effect on nontargeted fish as compared with trammel netting. Trammel netting in the riverine portion of the Colorado River downstream of Davis Dam, Arizona-Nevada yielded an average of 43 razorback suckers a year (1999 to 2005). Capture rates averaged 0.5 razorback suckers per staff day effort, at a cost exceeding $1,100 per fish. Population estimates calculated for 2003-2005 were 3,570 (95 percent confidence limits [CL] = 1,306i??i??i??-8,925), 1,768 (CL = 878-3,867) and 1,652 (CL = 706-5,164); wide confidence ranges reflect the small sample size. By-catch associated with trammel netting included common carp, game fish and, occasionally, shorebirds, waterfowl, and muskrats. Hoop nets were prone to downstream drift owing to design and anchoring problems aggravated by hydropower ramping. Tests were dropped after the 2006 field season and replaced with electrofishing. Electrofishing at night during low flow and when spawning razorback suckers moved to the shoreline proved extremely effective. In 2006 and 2007, 263 and 299 (respectively) razorback suckers were taken. Capture rates averaged 8.3 razorback suckers per staff day at a cost of $62 per fish. The adult population was estimated at 1,196 (925-1,546) fish. Compared with trammel netting, confidence limits narrowed substantially, from +or- 500 percent to +or- 30 percent, reflecting more precise estimates. By-catch was limited to two common carp. No recreational game fish, waterfowl, or mammals were captured or handled during use of electrofishing. Aerial photography (2006 and 2007) suggested an annual average of 580 fish detected on imagery. Identification of species was not possible; carp commonly have been mistaken for razorback sucker. Field verification determined that the proportion of razorback suckers to other fish was 3:1. On that basis, we estimated 435 razorback suckers were photographed, which equals 8.4 razorback suckers per staff day at a cost of $78 per fish. The data did not lend itself to population estimates. Fish were more easily identified from boats, where their lateral rather than their dorsal aspect is visible. On average, 888 razorback suckers were positively identified each year. Observation rates averaged 29.6 razorback suckers per staff day at a cost less than $18 per fish observed. Sucker densities averaged 20.5 and 9.6 fish/hectare which equated to an average spawning population at Needles, Calif., of 2,520 in 2006 and 1152 in 2007. The lower 2007 estimate reflected a refinement in sampling approach which removed a sampling bias. Electrofishing and boat surveys were more cost effective than other methods tested, and they provided more accurate information without the by-catch associated with trammel netting. However, they provided different types of data. Handling fish may be necessary for research purposes but unnecessary for general trend analysis. Electrofishing was extremely effective but can harm fish if not used with caution. Unnecessary electrofishing increases the likelihood of spinal damage and possible damage to eggs and potential young, and it may alter spawning behavior or duration. B