Aims Stable angina is the most common presentation of heart disease and has a good prognosis. With declining coronary artery disease (CAD), rates a diagnostic approach balancing costs and benefits is a challenge, particularly in women. This study describes the real-life diagnostic workup in a large hospital to explore whether the diagnostic approach may be improved. Methods and results We identified 4028 patients free of CAD, referred for and assessed with non-invasive (NIT) or invasive test for stable suspected CAD in 2012–15. In both the sexes, the majority (>85%) presented with chest pain as primary symptom. Women had more non-angina (60.2 vs. 54.5%) and less typical angina (8.2 vs. 11.8%, P < 0.001). Despite a mean pretest probability of 20.9% in women and 45.1% in men (P < 0.001), only 69 (3.1%) women and 190 men (10.4%) were diagnosed with obstructive CAD. In all, 93% underwent a NIT and 80% of these were normal. Among the 1238 men and 1595 women with non-angina or dyspnoea, only 6.1% and 2.9%, respectively, had positive NIT. After multiple adjustments, women remained less likely to have positive NIT [odds ratio (OR) 0.42 95% confidence interval (95% CI 0.32–0.56)] and given a positive test also less likely to have obstructive CAD [OR 0.30 (0.17–0.52)]. The C-statistics for predicting positive NIT was 0.77 (0.72–0.82) in women and 0.77 (0.74–0.80) in men. Conclusion These data confirm the very low diagnostic yield of non-invasive and invasive assessment of CAD in current clinical practice, particularly in women and in patients with atypical symptoms. Data call for a more rational approach to avoid unnecessary testing.