Study of over 1,000 specimens representing all notharctine genera and species leads to the conclusion that current concepts about the relationships of genera within the Notharctinae are incorrect. The following describes the more probable relationships among these genera. 1) Smilodectes and Notharctus are more closely related to each other than either is to any known early Eocene notharctine. Synapomorphies linking these genera include relatively narrow upper and lower molars, a relatively low-crowned P4, and paraconid size reduction on M1–3. 2) Among known Wasatchian notharctines, a clade consisting of Copelemur tutus and Cop. praetutus shares several lower molar synapomorphies with the Notharctus–Smilodectes clade, and therefore appears to form the Wasatchian sister group of Bridgerian notharctines. Synapomorphies documenting this relationship include well-developed entoconid notches on P4–M2, an anteriorly placed paraconid on M2, and a long premetacristid on M2. 3) Copelemur and Pelycodus are independently derived from early North American Cantius.
Recent suggestions that the European adapine taxa Leptadapis priscus and Microadapis sciureus share special phylogenetic relationships with Smilodectes are rejected. The reduced (or lack of a) paraconid and morphology of the paracristid and other features identified as synapomorphies linking these adapines with Smilodectes are also characteristic of most other adapines as well (e.g., other species of Leptadapis, Adapis, Europolemur, and Anchomomys). Such traits developed independently in Smilodectes, which is clearly a notharctine on the basis of many synapomorphies and thus are not evidence of a close phylogenetic relationship between Smilodectes and L. priscus or M. sciureus.