Prehistoric sites in Western Canada present unusual conditions for magnetic prospection procedures. However, by addressing specific types of features, particularly fire hearths and fired rock and pottery, useful information about the cultural content of an archaeological site can be obtained. To secure comparative data, a number of replicative studies were conducted, with specific reference to determining a typical anomaly produced by the features. A small fire pit was kindled several times and repeatedly monitored with surveys using a single proton magnetometer. Fired rock received similar treatment. An in situ collection of local aboriginal pottery was also assayed. The experimental results indicated that fired rock will produce a detectable magnetic field after one firing, but a hearth must be rekindled at least three times to produce a significant anomaly. Pottery fragments also generate a small magnetic field which requires that the material be very close to the site surface to be discerned. The experiments also suggested that interpretation is enhanced by obtaining two magnetic readings per sensor station and using their difference to minimize ambient field fluctuations and natural magnetic variation caused by subsurface geology. The model data were used to interpret the results of a magnetic assessment of a large prehistoric campsite in Saskatchewan. Excavation results agreed well with the information provided by the predictive models. Two temporary dwelling remnants and two pottery vessels were exposed in areas determined to be magnetically significant. The magnetic assessment technique, when used to locate specific feature types, can be useful in prehistoric archaeological site assessment. -- AATA