The coracoid pain test (CPT) could contribute to the diagnosis of frozen shoulder (FS) with palpation. However, due to assessor performance these values might be unreliable. Therefore, the aim was to explore the diagnostic accuracy of an instrument-assisted CPT and two alternative approaches (pain severity and side comparison) for assistance in the diagnosis of FS. Patients with FS and healthy age-matched controls were recruited. All participants underwent the instrument-assisted CPT on both shoulders with a pressure algometer. Sensitivity, specificity, and likelihood ratios were determined for the three approaches. In total, 35 patients with FS and 35 healthy participants were included. The original approach was positive in eight participants (11.4%), with only sufficient specificity to draw a conclusion. The pain severity approach was positive in 31 participants (44.3%) with sufficient sensitivity, specificity and likelihood ratios. The side comparison approach was positive in 10 participants (14.3%) with excellent specificity and positive likelihood ratio. The specificity of the instrument-assisted CPT can be used to increase the probability of FS with both the original and alternative approaches. Only the pain severity approach can draw a conclusion with a negative test result. This study should be repeated with a cross-sectional design to strengthen and confirm the conclusions.