Bmpr1a null and floxed alleles (Ahn et al. 2001; Mishina et al. 2002) were obtained on a mixed 129 and C57BL/6 background and maintained by random breeding. Mice carrying the null and floxed alleles were typically mated to Gdf5-Cre mice as shown in Figure 3. The resulting mice are on a mixed 129; C57Bl/6; FVB/N background, with both controls and mutant animals generated as littermates from the same matings. Whole-mount skeletal preparations were made from 34- to 36-d-old mice (Lufkin et al. 1992). Pairs of ears from euthanized 6-mo-old animals were removed, pinned, photographed, projected, and measured from the base of the curve formed between the tragus and antitragus to the farthest point at the edge of the pinnae. Grasping ability in 6-mo-old mice was measured by placing animals on a slender rod and timing how long they could remain suspended on the rod, to a maximum time allowed of 2 min. Data from five consecutive trials for each mouse were averaged. Range of motion assays were conducted on the MT/P1 and P1/P2 joints of the second hindlimb digit from euthanized 18-wk-old animals. Forceps were used to bend the joint to its natural stopping position, and the resulting angle was measured to the nearest 10° under 12.5× magnification using a 360° reticule. Analysis described in this section occurred on animals lacking R26R. Control mice included all nonmutant genotypes generated by Parent 1 being heterozygous for Gdf5-Cre and Bmpr1anull and Parent 2 being heterozygous for Bmpr1afloxP (see Figure 3). All statistical analysis used the Student's t-test or Welch's t-test, and values listed are mean ± standard error of the mean.