Background Osteoarthritis of the radiocarpal joints is commonly encountered by hand surgeons. To date, there is no well-defined method of radiographically grading osteoarthritis of the wrist. Methods Preoperative radiographs of 48 patients undergoing wrist arthroscopy were evaluated retrospectively. Images were graded subjectively by five surgeons based on overall severity of arthritis, osteophytes, subchondral cysts, and subchondral sclerosis. The joint space height (JSH) ratio was calculated by measuring the space of the mid-radioscaphoid and mid-radiolunate joints and dividing each by the height of the capitate. Arthroscopic grading of arthritis was obtained from operative records and compared to subjective and objective grades. ANOVA testing evaluated for statistical significance with p<0.05. Inter-rater and intra-rater reliability was determined using Pearson's correlation analysis and Cohen's kappa coefficient. Results Objective measurement using the JSH ratio demonstrated a significant decrease as arthroscopic arthritis grade increased for both radioscaphoid and radiolunate joints. Subjective grading of radioscaphoid and radiolunate joints was able to detect moderate/severe, but not mild arthritis. Subjective grading underestimated the degree of arthritis, particularly in the radiolunate joint. Inter-rater reliability was better for objective compared to subjective grading. Conclusions Subjective grading of wrist arthritis can detect moderate/severe radiocarpal arthritis but poorly evaluates early arthritis and underestimates severity. Objective grading using the JSH ratio accurately grades radioscaphoid arthritis and detects early radiolunate arthritis. The JSH ratio more accurately assesses radiocarpal arthritis compared with subjective grading. As there currently is no accepted method to radiographically grade wrist arthritis, the JSH ratio represents a promising option.