Circumferential remodeling of spiral arteries (SAs) during pregnancy is crucial for regulating maternal blood flow into the placenta and clinically important. However its mechanism is still ill defined in humans and mice. In mice, several important aspects of decidual SA remodeling (SAR) remain unexplored and were addressed here using morphometrics by examining SAs within the mesometrial half of the decidua basalis between embryonic day 6.5 (E6.5) and midgestation (E10.5). The data presented here provide evidence that supports the following novel conclusions about SAR: (a) SAs (defined by their muscular walls) appear between E6.5 and E7.5, undergo ‘outward hypertrophic’ SAR (SA lumen widening with muscular wall thickening) during the E7.5–E8.5 and E9.5–E10.5 periods, and ‘outward hypotrophic’ SAR (SA lumen widening with muscular wall thinning) during the E8.5–E9.5 interval. (b) ‘Outward hypotrophic’ SAR is associated with decreases in the overall number, but not density, of SA media nuclei, suggesting loss of SA muscular wall cells. Proximity of placenta-derived invasive trophoblast to SAs appears not be involved in SAR, as these cells were undetectable in the mesometrial region of decidua basalis throughout the E6.5–E10.5 period. Although the maternally derived lymphocytes of the decidual uterine natural killer (uNK) cell type are required for decidual SAR, the timing of this, the uNK cell parameter involved and the type of SAR they influence have not been adequately explored. Evidence is presented here that not all decidual SAR during this period is uNK cell-dependent. Rather, the data suggest that uNKs only influence ‘outward hypotrophic’ SAR during the E8.5–E9.5 period. Evidence is presented that the uNK cell parameter involved is the attainment of a certain maturation state (based on uNK cell size) by SA wall uNKs of the Dolichos biflorus agglutinin (DBA) lectin-positive uNK cell subset. This work also suggests that the previously shown loss of contractile mural cell character from the SA wall does not depend on either uNKs or trophoblast proximity. The novel implications of the present data for early mouse pregnancy are discussed.