In the mid-20th century, intracapsular femoral neck fractures were notorious as the 'unsolved fracture' due to high rates of non-union and osteonecrosis. Pauwels had classified fractures by the inclination angle of the fracture line, but this proved difficult to evaluate reliably on radiographs and did not account for degrees of rotational or translational displacement. Robert Symon Garden sought to provide a comprehensive, reproducible classification system based on fracture displacement and trabecular alignment, assessing how these patterns correlated with union rates and stability following low-angle nail fixation aligned with normal compressive trabecular lines.

Garden evaluated 80 consecutive patients with intracapsular femoral neck fractures treated with low-angle internal fixation (placing the implant more horizontally to parallel compressive trabecular load vectors). Fractures were assessed on anteroposterior and lateral radiographs and categorised into four distinct stages based on trabecular alignment and displacement. Patients were monitored clinically and radiographically for at least 12 months post-operatively to evaluate union, implant stability, and rates of avascular necrosis.

The Garden classification became the global reference standard for categorising intracapsular hip fractures and directing clinical decision-making. In contemporary practice, it is largely collapsed into a dichotomous framework: non-displaced (Garden I and II) versus displaced (Garden III and IV). This distinction underpins national guidelines (e.g., NICE, AAOS), where undisplaced fractures in elderly patients are often considered for internal fixation, whereas displaced fractures in physiologically older adults are predominantly treated with hemiarthroplasty or total hip arthroplasty to avoid the substantial reoperation risks associated with non-union and avascular necrosis.

Garden's original work provided seminal anatomical and biomechanical insights that transformed femoral neck fracture management. However, its primary modern critique is poor to moderate inter- and intra-observer reliability when distinguishing across all four individual stages on plain radiographs (particularly Type I vs II, and Type III vs IV). Furthermore, CT cross-sectional studies have demonstrated that true 'incomplete' Type I fractures are extremely rare and almost always complete fractures with posterior comminution and valgus/retroverted impaction. Nevertheless, reliability improves dramatically when simplified into a binary system (non-displaced vs displaced).