Abstract Background/Aims There are often significant delays to the diagnosis of axial spondyloarthritis (AxSpA). The imaging modality of choice to investigate for AxSpA is an MRI with a specific AxSpA protocol (MRI AS protocol). A common route of referral for MRI scans looking for AxSpA is via the community physiotherapy team. The Spondyloarthritis Diagnosis Evaluation (SPADE) tool has been designed to estimate the probability of patients having AxSpA. This study aimed to look at the relationship between SPADE tool probability score and MRI AS protocol results in patients who were referred from community physiotherapy. Methods This was a retrospective service evaluation. The sample was obtained from 200 MRI AS protocol requests between 6 January 2020 and 15 December 2021. For each patient, the MRI result was obtained from ICE electronic results software. MRI results were categorised as AxSpA, not AxSpA, or equivocal. For equivocal results, electronic patient notes were reviewed to identify whether patients went on to be diagnosed with AxSpA. The patient SPADE score (improbable, additional tests necessary, probable diagnosis, or definite diagnosis) was recorded for each patient. Information was obtained from the electronic physiotherapy notes to determine a SPADE score. Results Of the 200 scans requested, 21 scans were excluded (eight due to AS protocol not being performed; six due to duplicate requests; four not investigating for AxSpA; two where radiology had suggested MRI AS protocol; one patient awaiting recall for a repeat scan). Of the 179 patients who had MRI scans, 9.5% (n = 17) showed AxSpA, 8.9% were equivocal (n = 16) and 81.6% (n = 146) were negative. A SPADE score was determined for 177 patients (two had insufficient data to determine a score). 30% (n = 53) of patients were ‘improbable’, 56% (n = 99) were ‘additional tests necessary’, 8% (n = 14) were ‘probable’ and 6% (n = 11) were ‘definite’. Of the ‘improbable’ cases, three had MRI demonstrating AxSpA and one was equivocal (4/53 [7.5%]). 49 patients had no MRI evidence of AxSpA. Of the patients with MRI positive scans, 3/17 (17.6%) were ‘improbable’ using the SPADE tool. Conclusion Having a tool to screen patients in the community may help to better identify those who warrant further investigation for AxSpA. The SPADE tool may be a useful aid to identify who to refer for further investigation of inflammatory back pain. Many patients identified as ‘improbable’ by the SPADE tool did not have evidence of AxSpA. If the tool was applied prospectively, this may lead to large reduction in the number of MRI AS protocol scans performed. However, there were a small number of ‘improbable’ cases who had MRI evidence of AxSpA and could therefore be misdiagnosed as non-inflammatory back pain. Disclosure T. Nadin: None. S. Crompton: None. A. Chopra: None. N. Kotnis: None. L. Dunkley: Honoraria; Abbvie, UCB, Pfizer and Novartis. Other; Speakers fees from Abbvie, UCB, Pfizer and Novartis.