Primary implant stability (PIS) depends on surgical technique, implant design and recipient bone characteristics, among other factors. Bone density (BD) can be determined in Hounsfield units (HUs) using cone beam computed tomography (CBCT). Reliable prediction of PIS could guide treatment decisions. We assessed whether PIS was associated with recipient bone characteristics, namely, BD and alveolar ridge width (ARW), measured preoperatively by CBCT. We studied a convenience sample of 160 implants placed in 48 patients in 2016 and 2017. All underwent CBCT with a radiological/surgical guide yielding values for ARW and BD. PIS measures used were the implant stability quotient (ISQ) from resonance frequency analysis and insertion torque (IT). IT was most influenced by the HU value at 0.5 mm outside the implant placement area, followed by the value within this area and ISQ by the HU value at 0.5 mm outside the placement area, followed by implant placement site and apical ARW. ISQ values were significantly related to ARW in coronal (p <.05), middle (p <.01) and apical (p <.01) thirds. ISQs were higher with larger diameter implants (p <.01). ISQ and IT were strongly correlated (p <.001). PIS in terms of ISQ and IT is positively correlated with edentulous alveolar ridge BD measured by CBCT, implying that implant stability may be predicted preoperatively. Wide alveolar ridges favored lateral PIS but did not affect rotational PIS. The most significant predictor of lateral and rotational PIS in our patients was the HU value at 0.5 mm outside the implant placement area.