This report provides the results of a detailed Level II analysis of scour potential at structure RICHTH00030006 on Town Highway 3 crossing an unnamed tributary to the Missisquoi River, Richford, Vermont (figures 1–8). A Level II study is a basic engineering analysis of the site, including a quantitative analysis of stream stability and scour (U.S. Department of Transportation, 1993). Results of a Level I scour investigation also are included in Appendix E of this report. A Level I investigation provides a qualitative geomorphic characterization of the study site. Information on the bridge, gleaned from Vermont Agency of Transportation (VTAOT) files, was compiled prior to conducting Level I and Level II analyses and is found in Appendix D. The site is in the Green Mountain section of the New England physiographic province of northern Vermont. The 4.5-mi² drainage area is in a predominantly rural basin. In the vicinity of the study site, the surface cover is pasture upstream and downstream of the bridge. In the study area, the unnamed tributary to the Missisquoi River is a sinuous channel with a slope of approximately 0.008 ft/ft, an average channel top width of 39 ft and an average channel depth of 2 ft. The channel slope was obtained from a topographic map (USGS, 1986). The predominant channel bed material is gravel with a median grain size (D₅₀) of 26.2 mm (0.0861 ft). The geomorphic assessment at the time of the Level I and Level II site visit on June 28, 1995, indicated that the reach was stable. The Town Highway 3 crossing of an unnamed tributary to the Missisquoi River is a 26-ftlong, two-lane bridge consisting of one 24-foot concrete T-beam span (Vermont Agency of Transportation, written communication, March 9, 1995). The bridge is supported by vertical, concrete abutments with wingwalls. The channel is skewed approximately 40 degrees to the opening while the opening-skew-to-roadway is 0.0 degrees. The only scour protection measures at the site were type-2 stone fill (less than 36 inches diameter) along the upstream right wingwall and at the upstream end of the right abutment. Additional details describing conditions at the site are included in the Level II Summary and Appendices D and E. Scour depths and rock rip-rap sizes were computed using the general guidelines described in Hydraulic Engineering Circular 18 (Richardson and others, 1995). Total scour at a highway crossing is comprised of three components: 1) long-term streambed degradation; 2) contraction scour (due to accelerated flow caused by a reduction in flow area at a bridge) and; 3) local scour (caused by accelerated flow around piers and abutments). Total scour is the sum of the three components. Equations are available to compute depths for contraction and local scour and a summary of the results of these computations follows. Contraction scour for all modelled flows ranged from 1.7 to 1.8 ft. The worst-case contraction scour occurred at the 500-year discharge. Scour at the left abutment ranged from 7.6 to 12.6 ft with the worst case occurring at the 100-year event. Scour at the right abutment ranged from 1.6 to 5.6 ft with the worst case occurring at the 500-year event. Additional information on scour depths and depths to armoring are included in the section titled “Scour Results”. Scoured-streambed elevations, based on the calculated scour depths, are presented in tables 1 and 2. A cross-section of the scour computed at the bridge is presented in figure 8. Scour depths were calculated assuming an infinite depth of erosive material and a homogeneous particle-size distribution. It is generally accepted that the Froehlich equation (abutment scour) gives “excessively conservative estimates of scour depths” (Richardson and others, 1995, p. 47). Usually, computed scour depths are evaluated in combination with other information including (but not limited to) historical performance during flood events, the geomorphic stability assessment, existing scour protection measures, and the results of the hydraulic analyses. Therefore, scour depths adopted by VTAOT may differ from the computed values documented herein.