The purpose of this investigation was to determine the stability and extent of sexual, geographical, and individual variation both within and between three populations of the lizard Uta stansburiana stejnegeri and to assess evolutionary factors that might produce the observed interpopulational differences. This subspecies is abundant and easily collected in western Texas. Over a two-year period approximately 1900 individuals were taken from these populations and subjected to taxonomic analysis. Thirteen morphological characters were selected. Samples were rendered homogeneous with regard to sex and size groups. All data were treated statistically and significance tests applied to observed differences. Of the three populations studied, the one in Palo Duro 'Canyon, which is separated by a distance of 250 airline miles from either of the other two, is the most divergent. Gene flow between Palo Duro lizards and those to the south is more limited because of low lizard density in the unsuitable habitat in the region which separates them. Finally, for several reasons random genetic change may be of more importance in the northern population than in the southern ones. Although significant statistical differences have been established in a number of different characters, taxonomically all data fall below the conventional level for subspecific recognition (90% rule) of the populations. Some regional infrasubspecific differentiation due to local adaptation is indicated. INTRODUCTION Uta stansburiana ste jnegeri, the desert side-blotched lizard, is well suited for taxonomic study. It is abundant, easily collected in large samples throughout the year, and possesses scale counts and other meristic characteristics that exhibit a wide range of variation. For these reasons, this lizard was chosen for a study of microvariation at the intrasubspecific level. The objectives of the present study were to determine the extent of sexual, geographic and individual variation in three populations of Uta s. stejnegeri in western Texas on the basis of 13 selected morphological characters and to, attempt to determine what evolutionary factors were chiefly involved in producing the divergence in the three populations. Acknowledgments.-For assistance in the field, I am grateful to William Voss and Donald Woodard of Texas Technological College; and John George, University of Kansas. For supplying much helpful information I am especially indebted to William Hahn, Sumner Dana, and James Thomerson, all graduate students at this institution. Last, I am particularly grateful to my major professor, Dr. Donald W. Tinkle, for his patience, suggestions, and valuable assistance during the course of this study. A National Science Foundation Grant Number G-7143, and one from the U. S. Atomic Energy Commission Number