For long time the western-central Mexico has been affected by oblique subduction caused by Farallon plate beneath North America. As result, smaller plates (e.g. Cocos Plate), several fault systems outlining crustal blocks (e.g. Michoacán block) and magmatic arcs (e.g. Paleocene-Early Oligocene magmatism and the Trans-Mexican Volcanic Belt) were developed. Still, no paleomagnetic data are available for Oligocene and Miocene. The principal aim of this study is to evaluate whether the tectonic rotations and relative motions of these blocks occurred before the Miocene. Here, we report a detailed rock-magnetic and paleomagnetic results from Tecalitlan area, located in the Michoacán block. Sixteen sites (about 150 oriented samples) were collected including one radiometrically dated diabase dike (35.0 ± 1.8 Ma). Rock-magnetic experiments permitted identification of magnetic carriers and assessment of the paleomagnetic stability. Continuous susceptibility measurements vs temperature in most cases yield reasonably reversible curves with Curie points close to that of magnetite. Reliable paleomagnetic directions were obtained for 12 sites. Inclination I and declination D of the mean paleomagnetic direction obtained in this study are I = 33.1°, D = 345.0°, and Fisherian statistical parameters are k = 25, 95 = 8.9°. The corresponding mean paleomagnetic pole position is P lat = 75.7°, P long = 166.6°, K = 31, A 95 = 8.0°. The mean inclination is in Continuation of LATINMAG Special Issue #4, Stud. Geophys. Geod., 54 (2010) J. Rosas-Elguera et al. 266 Stud. Geophys. Geod., 55 (2011) reasonably good agreement with the expected value, as derived from reference poles for the stable North America. Magnetic declination is not significantly different from that expected which is in disagreement with a counterclockwise tectonic rotation of about 20º previously reported for the studied area. Based on paleomagnetic results obtained in this study compiled with those currently available from the Michoacán Block, we propose a simple model suggesting that sometime in Eocene epoch the convergence vector of the Farallon plate relative to North America plate was normal to the trench before reaching an actual oblique convergence.