This audit of 408 synovial fluid samples, analysed for cell counts and crys
tals, revealed that crystals were present in only 25 samples (6.1%). Of the
se, in only three patients was the diagnosis uncertain and therefore the an
alysis helpful. Cell counts and types generally reflected known underlying
diagnoses of inflammatory arthritis or osteoarthritis. Routine synovial flu
id analysis does not contribute to diagnosis or management in established r
heumatic disorders and should be performed only when the underlying cause i
s uncertain or in newly presenting patients. Major savings can be made by a
bandoning routine synovial fluid analysis.