Archaeological investigations in the Columbia Trench and the Rocky Mountains have occurred sporadically in the last fifty years.  Most of these have been focussed on assessments of development impacts.  In 1954, Charles Borden (1956) conducted the first archaeological surveys in the East Kootenay Region.  He recorded 25 sites in the Kootenay River valley north of the International Boundary and another 29 sites from Columbia Lake north to Windermere Lake.  Borden noted that his work was "intended to call attention to the archaeological resources and problems of the East Kootenay region and to prepare the way the future work" (1956:100). Among the issues that he identified was the relationship of the modern Kutenai and Kinbasket (Shuswap) to archaeological materials, the distribution of circular earth lodges and their association with winter dwellings of the Shuswap, and the categorization of projectile points into areal and temporal groups.

In 1975 and 1976, McKenzie test excavated two sites (EcQa 1 and EcPx 5) on the shores of Windermere Lake (McKenzie 1976).  Two undated cultural horizons were located at EcPx 5.  The upper horizon was characterized by small triangular side notched and corner notched projectile points, scrapers, a variety of other chipped stone tools, ground stone arrow shaft abraiders, a rib bone spatulate object, and chipped bones.  Projectile points with expanded stems and corner notches or broad side notches characterized the lower horizon plus a variety of other chipped stone tools.  Two short fragments of incised bird long bones were also recovered.  Most of the bone was heavily fragmented, burned or calcined.  Identifiable bone included deer phalanges and mandible fragments; as well as, portions of a right mandible of a mountain sheep (Ovis canadensis). A variety of lithic materials were utilized at this site.  In the upper horizon, variegated grey chert (opaque, 67.0%) and Top of the World chert (semi-opaque, 17.8%) were the most commonly utilized stone material types utilized for tools.  Variegated chert was the commonest lithic material found in the chipping debris (85%).  In the lower horizon, basalt and siltstone formed the largest percentage (71.8) of chipping detritus but a lower frequency (33.3 %) of stone tools.

Sneed (1979) conducted a detailed impact assessment of heritage resources of the proposed Kootenay River Diversion.  This spanned much of the upper Columbia River valley between Golden and Canal Flats and resulted in the recording of over 200 sites.  None of these sites were investigated in detail but the study indicated that there were significant archaeological resources in this area. 

A major excavation project on Columbia Lake at Site EbPw 1 was conducted by Mohs (1981) with additional excavation and synthesis conducted by Yip (1982). Results from this site indicated a diverse hunting-gathering-fishing subsistence base with seasonal occupations dating to the Late Prehistoric Period (A.D. 500 - 1800).  The site is large and appears to have occupations dispersed across the area. A large 9 m x 8 m circular house depression was test excavated.  Two circular depressions were also excavated which were likely roasting pits for processing vegetable foods. Projectile points recovered by Mohs (1981) included 28 side notched, 2 medial notched, 6 corner notched, 4 side indented and 1 contracting stem with rectangular shouldering.  Yip (1982) classified projectile points into two types:  small triangular shaped unnotched projectile points (10/21) and larger corner notched points (4/21).  Yip suggests that the corner-notched projectile points may have characterized the early part of the Late Prehistoric Period and that the small unnotched triangular points characterize the latter part of the Late Prehistoric Period.  Somewhat curious is the fact that Yip recovered no small side notched projectile points which formed 73% of the assemblage recovered by Mohs (1981).  This may be due to the researchers testing different areas within the site.  The different assemblages from the two areas may have resulted from occupations by different bands or groups either at the same time or at slightly different times.  An alternative explanation may be that the differences result from different activities that occurred in the two areas.  In both years a variety of other stone tools were also recovered including bifaces, unifaces, and a graver.  Bone artifacts included a bone bipoint and tip sections from four other pointed bone objects (Yip 1982), a tooth pendant and fragments of 18 ground or polished bone tools. Identified faunal remains included deer, caribou, rabbit, mink, weasel, martin, ground squirrel and beaver.  Fish remains identified included salmon, sucker and peamouth.  A small number of unidentified bird bones were also recovered (Yip 1982).

Just downstream of the north end of Columbia Lake is the area of Fairmont Hot Springs.  Several pit houses have been recorded in this area (Choquette 1971). This area has been largely destroyed by construction of a golf course and resort development.  Site EbPv 3 consisted of 10 pit house depressions and 3 cache pits located on a series of river terraces on the east bank of the Columbia River. At a second site, EbPv 14, four other pithouses were recorded.  In addition, considerable lithic materials, bone and fire broken rock was noted or collected.  There were considerable differences between the pithouse depressions at the two sites.  At EbPv 3, the pits were "saucer-shaped" being fairly shallow with diameters ranging from 2.55 to 5.32 m and averaging 3.6 m.  These ranged in depth from 0.12 to 0.37 m and averaged .21 m.  At EbPv 14 the six depressions were "bowl-shaped" and ranged in diameter from 3.0 to 6.1 m and averaged 4.4 m.  These ranged in depth from 0.53 to 0.82 m and averaged 0.70 m. One corner notched projectile point was recovered from each of EbPv 3 and EbPv 14.  Richards and Rousseau (1987) provide data on housepit sizes in the Canadian Plateau centered on Fraser and Thompson rivers.  They note that the smallest housepits there occur during the Plateau Horizon where they average 6.14 m in diameter.  These are associated with a variety of corner notched projectile points and date between approximately 2400 and 1200 BP.  The pithouse depressions at Fairmont Hot Springs may also date to this period.   Pithouse depressions have been recorded northward along the Upper Columbia (Sneed 1979) as well as in the upper Bow  and upper Red Deer River in Banff National Park. 

Eight housepit sites are known from Banff National Park (Langemann 1995, 1998).  The Banff sites have yielded radiocarbon dates between 2800 and 440 BP.   Most are between 2.5 and 3.5 m in diameter but some may be up to 5.0 m.  At the Drummond Glacier site, in the upper Red Deer Valley, the depressions are all between 2.8 and 3.4 m in diameter. Stone tools recovered from these sites span a wide time range.  Projectile points identified at these sites (although not necessarily from the depressions themselves) include: Besant and Pelican Lake Phase (Drummond Glacier Site); Agate Basin, Bitterroot side notched, McKean, similar to Late Prehistoric points from Arrow Lakes area, Timber Ridge side notched/Kamloops phase (Spring Site), Shuswap Phase, Second Lake Phase, and Pelican Lake (Divide Creek Site) (Greaves 1998).  Such a range of variation may be due to repeated occupations at these sites so that some of these may not be directly related to the house pits.  At these sites on the eastern side of the Rocky Mountains there appears to be associations to the culture areas both to the east and the west, however this has not been clearly defined.  There continues to be some confusion attributable to the overlap in characteristics of projectile point styles from the plains and the plateau during some time periods, especially in the Late Prehistoric Period.  Perhaps some of this confusion will be clarified in the future through more detailed study.  Lithic material types may be significant in this regard.  Most of the stone tools are manufactured on local materials but some items are made of distinctive materials such as Banff Chert, Norquay chert, and Top of the World chert so that their distributions might be indicative of transport or trade.

Another significant archaeological investigation was conducted at Site EdQa 8 (Bussey 1986).  Salvage excavations were conducted due to road revisions on a high terrace overlooking Windermere Lake.  The site has been assigned to the Late Middle Prehistoric Period (2000 B.C. to A.D. 400) with dates of 2130 +/- 70 B.P. and 2360 +/- 80 B.P.    The site likely functioned as a temporary camp and processing site.  Identifiable mammal remains include elk, wolverine, deer, lynx and beaver.  Some identified bird remains were also recovered.  Fish vertebrae were sucker.  Only two complete projectile points were recovered.  These were corner notched with expanding stems similar to Pelican Lake type points found in the Plains and Rocky Mountains at a similar time period. Top of the World chert and pale cream amber/amber chalcedony were the dominant lithic materials.

The materials recovered from these few sites sampled in the upper Columbia River area are inadequate to establish a detailed culture history.  Few of these sites are well dated and none date from the Early Prehistoric or Early Middle Prehistoric Periods.  Only Site EdQa 8 dates from the Late Middle Prehistoric Period (2000 BC to AD 400).  The remaining sites (EcPx 5, and EbPw 1) date from the Late Prehistoric Period (AD 400 (?) to AD 1850).   

Just to the east of the Columbia Trench is Kootenay National Park where Heitzmann (1997, 1998, 1999) has conducted test excavations at several sites.  All of these were small transitory hunting camps occupied for brief lengths of time.   Site 430T in Sinclair Col is a high elevation (7200 ft, 2100m) grassy bowl just below the Sinclair Kindersley pass.   A sample of animal bone fragments yielded a date of 1760 +/- 190 years B.P. (Heitzmann 1997).  Site 494T, along the Kootenay River has Late Prehistoric materials recovered including a side notched projectile point made of Top of the World chert and other stone materials.  Blood residue on a stone tool has been identified as bos/bison.  As this site is radiocarbon dated to 380 +/- 50 years B.P. it has been suggested that bison was hunted at this site.  Site 497T is located on a high terrace of the Kootenay River and yield a projectile point with broad deep corner notches and straight base made of black siliceous siltstone.  The point is stylistically similar to one located at the Lehman Site near Kamloops, assigned to the Nesikep Tradition of the Early Middle Prehistoric Period (6000 to 4500 years BP).  Similar projectile points have also been found along the Pend D'Oreille River near Nelson and are suggested to date to the Upper Middle Period (1500 B.C. to A.D. l) (Bussey 1981:47).  Also in Kootenay National Park, Heitzmann (1995) tested a quartz crystal workshop at Kaufmann Lake.  Carbon from a hearth at this site yielded a date of 4, 470 years B.P. +/- 80 years (BGS 1774).  These sites indicate that hunting and lithic extraction were occasional activities that occurred in the more isolated mountain regions east of the Columbia Trench.

Further south Blake (1981) tested at a site adjacent to the Wild Horse River near Fort Steele.  No time diagnostic materials, other than historic artifacts,  were recovered and no dates are available for this site.  However "about half the total sample of flakes were of Top of the World chert.  Faunal species identified included mule deer (Odocoileus heminous), wapiti (Cervus elaphus), cow/bison,  lynx (Lynx rufus), dog, microtus, bird and fish.  One of the elements was confidently identified as a distal left radius of a bison.  The remaining cow/bison elements may be either species.