Small volcanic edifices, shields with a diameter less than about 20 km, are common and sometimes very abundant features on the plains of Venus. Typically, they form tight or loose clusters of features known as shield fields. Small shields are interpreted to be formed due to small-scale eruptions through numerous and distinct sources, a mode of formation apparently significantly different from the mechanism thought to be responsible for the emplacement of the vast regional plains of Venus. Did the eruption style of small shields occur repeatedly throughout the visible part of the geologic record of Venus? Or was this style more concentrated in a specific epoch or epochs of geologic history? Do the clusters of shields represent localized development of sources over a thermal anomaly such as a plume, or do they represent exposures or kipukas of a more regional unit or units? A major step toward answering these questions is an understanding of small shield stratigraphy. Multiple criteria have been developed to assess the stratigraphic relationships of individual small shields and that of shield fields with the adjacent units. In our analysis, we expanded and developed the previous criteria and added detailed criteria to describe specific patterns of deformation within shield fields, cross-cutting, and embayment relationships between shield fields and surrounding units. We also used secondary characteristics of shield fields such as radar albedo difference, changes in shield density and size, etc. In our study, we applied these criteria and analyzed in detail stratigraphic relationships of shield fields in a random sample of features (64 fields) and in the global geotraverse along 30¿N (77 fields). The total number of analyzed shield fields (141) represents about 22% of the general population of these features catalogued by Crumpler and Aubele <2000>. The majority of the fields (98, or ~69%) predate emplacement of material of vast regional plains with wrinkle ridges. Fifteen fields (~11%) appear to be synchronous with regional plains, and eleven fields (~8%) postdate the plains. Nine fields (~6%) display ambiguous relationships with regional plains and their relative age is uncertain, and eight fields (~6%) represent unclear cases when fields are covered by crater-related materials or by young lava flows or are not in contact with regional plains. The results of our study provide evidence for a distinct change of volcanic style from the mode of formation of globally abundant small shields to the mode of emplacement of vast regional plains in many areas on Venus. This systematic change of volcanic style appears to be inconsistent with the nondirectional or quasi steady state character of the geologic record of Venus. Although individual small shields were formed throughout themajority of the visible geologic history of Venus, in the syn- and postregional plains time the small-shields style of volcanism was significantly reduced in abundance. The shield fields that predate regional plains do not display a strong tendency to form a single group or a few groups and can be found virtually in all places on Venus. We interpret this observation to mean that these shield fields were globally distributed before the emplacement of regional plains. This interpretation means that the shield fields embayed by regional plains represent exposures of a specific, globally widespread unit, shield plains (psh). In contrast, shield fields that postdate regional plains occur preferentially in the Beta-Atla-Themis region on Venus, well known for its concentration of relatively young volcanic and tectonic activity. The spatial association of relatively young fields with the large centers of young volcanism suggests a genetic link of these fields with the formation and development of the large-scale volcanic centers. The abrupt decrease of the number of shields that postdate the formation of shield plains (psh) strongly suggests a major change of the style of volcanism following their emplacement as a globally distributed unit.